* [[https://www.gnu.org/savannah-checkouts/gnu/bash/manual/bash.html|bash reference manual - gnu.org]] * [[http://www.kornshell.com/doc/faq.html|Lots of interesting facts about Korn shell ksh93 (including a shell compiler shcomp) - kornshell.com]] * [[http://mywiki.wooledge.org/BashPitfalls#for_f_in_.24.28ls_.2A.mp3.29|Bash Pitfalls]] - like why does this for f in $(ls *.mp3) not work?!! * [[http://hyperpolyglot.org/unix-shells#var-expansion|Good shell comparison including variable expansion - hyperpolyglot.org]] * [[http://pubs.opengroup.org/onlinepubs/9699919799/utilities/toc.html|Shell coding standards - opengroup.org]] * [[http://aix4admins.blogspot.be|aix4admins]] * [[http://linuxgazette.net/issue55/okopnik.html|The Deep, Dark Secrets of Bash]] * [[http://www.speakeasy.org/~arkay/216-7.4KshFunctions.html|Korn Shell Functions]] * [[http://www.grymoire.com/Unix/AwkRef.html|AWK reference from Bruce Barnett]] * [[http://www.catonmat.net/blog/wp-content/uploads/2008/09/awk1line.txt|HANDY AWK ONE-LINERS]] * [[http://www.grymoire.com/unix/Quote.html|Unix quoting from from Bruce Barnett (grymoire)]] * [[http://www.mpi-inf.mpg.de/departments/rg1/teaching/unixffb-ss98/quoting-guide.html|Another excellent quoting tutorial]] * [[http://aix4admins.blogspot.be/2015/02/snapshot-fs-snapshot-snapshot-is-state.html|AIX snapshot technology]] * [[RaspberryPi]] * [[SSH]] * [[http://www.anattatechnologies.com/q/2013/04/multiuser-screen/|Screen sharing from Unix using screen]] * [[https://robots.thoughtbot.com/a-tmux-crash-course|Screen sharing from Unix using tmux]] * [[http://www.csb.yale.edu/userguides/wordprocess/vi-summary.html|Handy ex and vi reference]] * [[Fedora install for my own purposes]] * [[http://www-01.ibm.com/support/docview.wss?uid=swg21282999|UNIX commands that can be used for performance-related issues]] * [[http://www.billharlan.com/papers/Bourne_shell_idioms.html|Bourne shell idioms - billharlan.com]] ==== Find all database-related files from the controlfile ==== Assumes a backup to trace of the controlfile is available perl -ne 'print unless 1../Set #2. RESETLOGS/' "${CTRLFILE}"|grep -v SETCONFIG|grep -E "\'.+\/"|awk -F"'" '{print "rm -f "$2}' ==== List files using find without recursively descending into subdirectories ==== * [[https://unix.stackexchange.com/questions/786454/depth-1-vs-mindepth-1-maxdepth-1|Stack Exchange]] The idiomatic way is find . ! -name . -prune -name "filestosearch*" The key is the placement on the line of the -prune parameter set lines 1000 col stmt for a500 select 'find '||destination||' -mtime +28 -prune -type f -ls' stmt from v$archive_dest where destination is not null; select 'find '||destination||' -mtime +28 -prune -type f -exec rm {} \;' stmt from v$archive_dest where destination is not null; This searches for and deletes all files in the archivelog destination older than 28 days without descending into the FRA subdirectory! It does no harm as the backup retention is only guaranteed for 28 days anyway. So all this does is clean up files that may have been lying around after a clone startup or before a db was switched to noarchivelog mode. ==== Print shell commands as they are being run ==== * [[https://stackoverflow.com/questions/71117953/how-to-write-bash-function-to-print-and-run-command-when-the-command-has-argumen|tackoverflow]] Note: Support for local - was added in Bash 4.4. function print_and_run_cmd { local PS4='Running cmd: ' local - set -o xtrace "$@" } ==== How to assign a certain keyboard shortcut to paste specific item ==== If installed, xsel and xvkbd can be used xsel -ib <<< 'Your string goes here' xsel -ib < your-file.txt xvkbd -file - <<< 'Your string goes here' xvkbd -file your-file.txt Assigning it to a Key combination in the window manager might require enclosing everything in quotes bash -c "xsel -ib <<< 'my string'" ==== A poor man's 'expect' ==== * [[https://www.shellscript.sh/hints.html|Logging into, and using, telnet in batch mode, having the commands piped to it from a shell script]] Or, using a shell's own logfile to interrogate what it is, itself, doing! Use something like this ./telnet_prompts.ksh | telnet >telnet.log #!ksh host="127.0.0.1" port=23 login="someuser" passwd="somepass" cmd="ls" timeout=3 file="telnet.log" prompt="$" echo open ${host} ${port} sleep 1 tout=${timeout} while [ "${tout}" -ge 0 ] do if tail -1 "${file}" 2>/dev/null | egrep -e "login:" > /dev/null then echo "${login}" sleep 1 tout=-5 continue else sleep 1 tout=`expr ${tout} - 1` fi done if [ "${tout}" -ne "-5" ]; then exit 1 fi tout=${timeout} while [ "${tout}" -ge 0 ] do if tail -1 "${file}" 2>/dev/null | egrep -e "Password:" > /dev/null then echo "${passwd}" sleep 1 tout=-5 continue else if tail -1 "${file}" 2>/dev/null | egrep -e "${prompt}" > /dev/null then tout=-5 else sleep 1 tout=`expr ${tout} - 1` fi fi done if [ "${tout}" -ne "-5" ]; then exit 1 fi > ${file} echo ${cmd} sleep 1 echo exit ==== Remove a directory from the PATH variable ==== Change all the colons into carriage returns, exclude the lines containing the directory and rejoin the lines with a colon! export PATH=$(echo $PATH | tr ":" "\n" | grep -v '/usr/local/bin' | xargs | tr ' ' ':') ==== Random password generator for SQL*Plus or PL/SQL ==== select dbms_random.string('a',15)||'#'||dbms_random.string('x',16) from dual; ==== Inventory commands for AIX ==== From https://www.netiq.com/documentation/platespin-recon-42/recon_user/data/b17df8u6.html Command Description rpm Get a list of installed programs. ps Get service (daemon) information. mount Get list of mounted volumes. df Get size information for mounted volumes. lslv Get SerialNumber and Label for mounted volumes. hostname Get hostname. namerslv Get DNS domain name. getconf Determine whether a 64-bit kernel. uname Determine OS release version. ifconfig Get configured network adapters and their IP address. netstat Get MAC address. entstat Get speed of network adapter. prtconf Get processor information and default gateway. lspv Get list of disks and their partitions. bootinfo Get disk size. lscfg Get disk model, manufacturer and type and network adapter model. lsitab Get daemon information. getconf Get machine memory and bootloader information. lsattr Get machine serial number and model,hyperthreading details and subnet mask. ==== Monitoring commands for AIX ==== Command Description vmstat Get Available Memory in Bytes ifconfig Get NIC information entstat Get Network Packets, Bytes Send/Received mpstat Get Percentage processor time iostat Get Disk transfer per second, disk idle, disk byte per second sar Get Disk Queue length, Percentage processor used lspv Get Disk partitioned or Unpartitioned space, disk list bootinfo Get disk size df Get the volume level information oslevel Get the version of AIX ==== AIX hacmp (PowerHA) startup and shutdown scripts ==== /usr/es/sbin/cluster/utilities/cllsserv -h -c gives something like #Name:Start_script:Stop_script:start_mode:cpu_usage_monitor:process_to_monitor_cpu:cpu_usage_monitor_interval: hn481_app:/ux/hacmp/bin/app_start.ksh HN481:/ux/hacmp/bin/app_stop.ksh HN481:background:no::10: === The app_start.sh script referenced above === #!/usr/bin/ksh #----------------------------------------------------------------- # # start_applications # #----------------------------------------------------------------- # # Customer : CLN # Environment : ALL # version : 3.0 # date : 11/2008 # syntax : app_start.ksh # description : This script is called by HACMP to start the HA # applications. It reads ../hacmp/rc2.d directory # and runs all script beginning whith S* # remarks : # history : # #================================================================== FILENAME=`basename $0` if [ $# -ne 1 ] then echo "$FILENAME require a parameter" exit 1 fi NODE=$1 if [ ! -d /ux/hacmp/$NODE ] then echo "!!! No /ux/hacmp/$NODE directory available !!!" exit 1 fi export HACMP_RG_NAME=$NODE APPBIN=/ux/hacmp/$NODE/bin APPLOG=/uxlog/hacmp/$NODE APPFIL=/ux/hacmp/$NODE/file APPTMP=/ux/hacmp/$NODE/tmp APPRC2=/ux/hacmp/$NODE/rc2.d if [ ! -d $APPLOG ] then for dir in app oracle samba do mkdir -p $APPLOG/$dir done fi LOGFILE="$APPLOG/$FILENAME.`date +%d%m%y`.`date +%H%M`.log" exec 1>$LOGFILE 2>&1 DATE=`date '+%Y-%m-%d_%H:%M:%S'` echo '\n###' echo "### Start of $0 - $DATE" for f in $APPRC2/S* { if [ -s ${f} ] then DATE=`date '+%Y-%m-%d_%H:%M:%S'` echo '\n###' echo "### Start of $f - $DATE" echo '###\n' /usr/bin/ksh ${f} start rc=$? DATE=`date '+%Y-%m-%d_%H:%M:%S'` echo '\n###' echo "### End of $f - RC=$rc - $DATE" echo '###\n' fi } DATE=`date '+%Y-%m-%d_%H:%M:%S'` echo '\n###' echo "### End of $0 - $DATE" exec 1<&- exit 0 ==== Other userful cluster commands ==== /usr/es/sbin/cluster/utilities/cllsres APPLICATIONS="hn481_app" FILESYSTEM="" FORCED_VARYON="false" FSCHECK_TOOL="fsck" FS_BEFORE_IPADDR="true" RECOVERY_METHOD="sequential" SERVICE_LABEL="hn481" SSA_DISK_FENCING="false" VG_AUTO_IMPORT="false" VOLUME_GROUP="ora1dwhtvg ora1prgvg app1prgvg app1tstvg ora1tstvg ora1tstarchvg hn481vg" USERDEFINED_RESOURCES="" /usr/es/sbin/cluster/events/cmd/cllsev -h -c #name:cmd:desc acquire_service_addr:/usr/es/sbin/cluster/events/acquire_service_addr:Script run when a service address is acquired on a node. acquire_takeover_addr:/usr/es/sbin/cluster/events/acquire_takeover_addr:Script run when a service address is acquired on a node. admin_op:/usr/es/sbin/cluster/events/admin_op:Script run when an administrator initiates a cluster operation. cluster_ffdc:/usr/es/sbin/cluster/events/cl_ffdc:Script run to collect first failure data capture cluster_notify:/usr/es/sbin/cluster/events/cluster_notify:Script run when SystemMirror detects a cluster configuration problem config_too_long:/usr/es/sbin/cluster/events/config_too_long:Script run when a cluster event has been running for a long time. event_error:/usr/es/sbin/cluster/events/event_error:Script run when there is a unrecoverable error while running a cluster event. external_resource_state_change:/usr/es/sbin/cluster/events/external_resource_state_change:Scripts run when a cluster resource encounters a recoverable error external_resource_state_change_complete:/usr/es/sbin/cluster/events/external_resource_state_change_complete:Scripts run when a cluster resource encounters a recoverable error fail_interface:/usr/es/sbin/cluster/events/fail_interface:Script run after a network interface has failed. forced_down_too_long:/usr/es/sbin/cluster/events/forced_down_too_long:Script run when a cluster node has been in the unmanaged state for some time get_disk_vg_fs:/usr/es/sbin/cluster/events/get_disk_vg_fs:Script run to acquire disks, varyon volume groups, and mount filesystems. intersite_fallover_prevented:/usr/es/sbin/cluster/events/intersite_fallover_prevented:Script run when a cluster resource is not moved across a site boundary join_interface:/usr/es/sbin/cluster/events/join_interface:Script run after a network interface has recovered. network_down:/usr/es/sbin/cluster/events/network_down:Script run when a network has failed. network_down_complete:/usr/es/sbin/cluster/events/network_down_complete:Script run after the network_down script has successfully completed. network_stable:/usr/es/sbin/cluster/events/network_stable:Script run when a network is stable. network_unstable:/usr/es/sbin/cluster/events/network_unstable:Script run when a network is continuously changing state. network_up:/usr/es/sbin/cluster/events/network_up:Script run after a network has become active. network_up_complete:/usr/es/sbin/cluster/events/network_up_complete:Script run after the network_up script has successfully completed. node_down:/usr/es/sbin/cluster/events/node_down:Script run when cluster services are stopped or a node has failed. node_down_complete:/usr/es/sbin/cluster/events/node_down_complete:Script run after the node_down script has successfully completed. node_down_local:/usr/es/sbin/cluster/events/node_down_local:Script run when it is the local node which is leaving the cluster. node_down_local_complete:/usr/es/sbin/cluster/events/node_down_local_complete:Script run after the node_down_local script has successfully completed. node_down_remote:/usr/es/sbin/cluster/events/node_down_remote:Script run when it is a remote node which is leaving the cluster. node_down_remote_complete:/usr/es/sbin/cluster/events/node_down_remote_complete:Script run after the node_down_remote script has successfully completed. node_up:/usr/es/sbin/cluster/events/node_up:Script run when a node is attempting to join the cluster. node_up_complete:/usr/es/sbin/cluster/events/node_up_complete:Script run after the node_up script has successfully completed. node_up_local:/usr/es/sbin/cluster/events/node_up_local:Script run when it is the local node which is joining the cluster. node_up_local_complete:/usr/es/sbin/cluster/events/node_up_local_complete:Script run after the node_up_local script has successfully completed. node_up_remote:/usr/es/sbin/cluster/events/node_up_remote:Script run when it is a remote node which is joining the cluster. node_up_remote_complete:/usr/es/sbin/cluster/events/node_up_remote_complete:Script run after the node_up_remote script has successfully completed. reconfig_configuration_complete:/usr/es/sbin/cluster/events/reconfig_configuration_complete:Script run while processing a dynamic configuration change. reconfig_resource_acquire:/usr/es/sbin/cluster/events/reconfig_resource_acquire:Script run while processing a dynamic configuration change. reconfig_resource_acquire_fence:/usr/es/sbin/cluster/events/reconfig_resource_acquire_fence:Script run while processing a dynamic configuration change. reconfig_resource_acquire_secondary:/usr/es/sbin/cluster/events/reconfig_resource_acquire_secondary:Script run while processing a dynamic configuration change. reconfig_resource_complete:/usr/es/sbin/cluster/events/reconfig_resource_complete:Script run while processing a dynamic configuration change. reconfig_resource_complete_secondary:/usr/es/sbin/cluster/events/reconfig_resource_complete_secondary:Script run while processing a dynamic configuration change. reconfig_resource_release:/usr/es/sbin/cluster/events/reconfig_resource_release:Script run while processing a dynamic configuration change. reconfig_resource_release_fence:/usr/es/sbin/cluster/events/reconfig_resource_release_fence:Script run while processing a dynamic configuration change. reconfig_resource_release_primary:/usr/es/sbin/cluster/events/reconfig_resource_release_primary:Script run while processing a dynamic configuration change. reconfig_resource_release_secondary:/usr/es/sbin/cluster/events/reconfig_resource_release_secondary:Script run while processing a dynamic configuration change. reconfig_topology_complete:/usr/es/sbin/cluster/events/reconfig_topology_complete:Script run while processing a dynamic configuration change. reconfig_topology_start:/usr/es/sbin/cluster/events/reconfig_topology_start:Script run while processing a dynamic configuration change. release_service_addr:/usr/es/sbin/cluster/events/release_service_addr:Script run when a service address is released from a node. release_takeover_addr:/usr/es/sbin/cluster/events/release_takeover_addr:Script run when a service address is released from a node. release_vg_fs:/usr/es/sbin/cluster/events/release_vg_fs:Script run to unmount filesystems and varyoff volume groups. rep_disk_notify:/usr/es/sbin/cluster/events/rep_disk_notify:Script run when a repository disk has failed resource_group_in_error:/usr/es/sbin/cluster/events/resource_group_in_error:Script run when a resource group goes to error state resource_state_change:/usr/es/sbin/cluster/events/resource_state_change:Scripts run when a cluster resource encounters a recoverable error or the user moves resource groups. resource_state_change_complete:/usr/es/sbin/cluster/events/resource_state_change_complete:Scripts run when a cluster resource encounters a recoverable error or the user moves resource groups. rg_move:/usr/es/sbin/cluster/events/rg_move:Scripts run when SystemMirror moves resources and resource groups. rg_move_acquire:/usr/es/sbin/cluster/events/rg_move_acquire:Scripts run when SystemMirror moves resources and resource groups. rg_move_complete:/usr/es/sbin/cluster/events/rg_move_complete:Scripts run when SystemMirror moves resources and resource groups. rg_move_fence:/usr/es/sbin/cluster/events/rg_move_fence:Scripts run when SystemMirror moves resources and resource groups. rg_move_release:/usr/es/sbin/cluster/events/rg_move_release:Scripts run when SystemMirror moves resources and resource groups. server_down:/usr/es/sbin/cluster/events/server_down:Script run when an application has failed. server_down_complete:/usr/es/sbin/cluster/events/server_down_complete:Script run when an application has failed. server_restart:/usr/es/sbin/cluster/events/server_restart:Script run to restart an application. server_restart_complete:/usr/es/sbin/cluster/events/server_restart_complete:Script run to restart an application. site_down:/usr/es/sbin/cluster/events/site_down:Script run when all nodes in a site are down. site_down_complete:/usr/es/sbin/cluster/events/site_down_complete:Script run after the site_down script has successfully completed. site_down_local:/usr/es/sbin/cluster/events/site_down_local:Script run when it is the local site which is leaving the cluster. site_down_local_complete:/usr/es/sbin/cluster/events/site_down_local_complete:Script run after the site_down_local script has successfully completed. site_down_remote:/usr/es/sbin/cluster/events/site_down_remote:Script run when it is a remote site which is leaving the cluster. site_down_remote_complete:/usr/es/sbin/cluster/events/site_down_remote_complete:Script run after the site_down_remote script has successfully completed. site_isolation:/usr/es/sbin/cluster/events/site_isolation:Scripts run when SystemMirror detects a site isolation or merge condition. site_isolation_complete:/usr/es/sbin/cluster/events/site_isolation_complete:Script run after a site_isolation event completes. site_merge:/usr/es/sbin/cluster/events/site_merge:Scripts run when SystemMirror detects a site isolation or merge condition. site_merge_complete:/usr/es/sbin/cluster/events/site_merge_complete:Script run after a site_merge event completes. site_up:/usr/es/sbin/cluster/events/site_up:Script run when cluster services are started on the first node in a site. site_up_complete:/usr/es/sbin/cluster/events/site_up_complete:Script run after the site_up script has successfully completed. site_up_local:/usr/es/sbin/cluster/events/site_up_local:Script run when it is the local site which is joining the cluster. site_up_local_complete:/usr/es/sbin/cluster/events/site_up_local_complete:Script run after the site_up_local script has successfully completed. site_up_remote:/usr/es/sbin/cluster/events/site_up_remote:Script run when it is a remote site which is joining the cluster. site_up_remote_complete:/usr/es/sbin/cluster/events/site_up_remote_complete:Script run after the site_up_remote script has successfully completed. split_merge_prompt:/usr/es/sbin/cluster/utilities/cl_sm_prompt:Script to prompt the operator for manual choice on split or merge start_server:/usr/es/sbin/cluster/events/start_server:Script run to start an application. start_udresource:/usr/es/sbin/cluster/events/start_udresource:Script run to start application servers. stop_server:/usr/es/sbin/cluster/events/stop_server:Script run to start an application. stop_udresource:/usr/es/sbin/cluster/events/stop_udresource:Script run to stop application servers. swap_adapter:/usr/es/sbin/cluster/events/swap_adapter:Script run to swap IP Addresses between two network interfaces. swap_adapter_complete:/usr/es/sbin/cluster/events/swap_adapter_complete:Script run after the swap_adapter script has successfully completed ==== cat a local file to the end of a file on a remote host over ssh ==== Nice trick but not really a trick! It saves having to scp the file over and then login and cat it on the remote host. The contents of the local file are catted to STDOUT which is piped over ssh to a waiting cat which appends the incoming STDIN to the specified file. cat id_rsa.pub | ssh oracle@hn5306 'cat >> .ssh/authorized_keys' ==== Countdown timer loop ==== Two examples from /usr/es/sbin/cluster/etc/rc.cluster #!/bin/ksh93 typeset i=3 # make sure init has time to start it while (( i-- )) do src_running=$(ps -e | awk '$NF == "srcmstr" { print $1; exit }') [[ -n "$src_running" ]] && break done if (( $i == 0 )) then cl_echo 234 "\n$PROGNAME: srcmstr is not running.\n" $PROGNAME exit 1 fi if [[ -n "$src_running" ]] then cl_echo 235 "Checking for srcmstr active...\n" typeset -i i=10 # try ten times to contact it while (( i-- )) do lssrc -s inetd >/dev/null 2>&1 && break # break out on success sleep 1 # otherwise wait a second and try again echo ".\c" done if (( i == 0 )) then cl_echo 236 "\n\nERROR: srcmstr is not accepting connections\n" exit 1 fi cl_echo 237 "complete.\n" fi ==== Sort a comma separated list in shell ==== * [[https://stackoverflow.com/questions/47665195/how-to-sort-comma-separated-values-in-bash|How to sort comma separated values in shell]] Not as easy as it sounds. Googling it shows many varied ways mostly using ugly label syntax.\\ In amongst all the chaff, I spotted this wheat! echo "7, 15, 6, 2, -9" | sed -e $'s/,/\n/g' | sort -n | tr '\n' ',' | sed 's/,$//' sed -e $'s/,/\n/g' : For splitting string into lines by comma. sort -n : Then you can use sort by number tr '\n' ',' : Convert newline separator back to comma. sed 's/,$//' : Removing trailing comma. It's that dollar sign that makes it possible. Without it, it fails. I have no idea what it does and I can find no info on it! ==== Generate a random sequence of characters (maybe for a password) ==== echo $(< /dev/urandom tr -dc %A-Z-a-z-0-9 | head -c${1:-30}) As that one might generate a number as the first character, this one generates 7 characters, followed by a '#', followed by 8 numbers. Should satisfy most ORAPW=$(echo "$(tr -dc A-Za-z ==== How to automatically download and process email attachments from Gmail ==== === Useful references === * [[http://www.courier-mta.org/maildrop/maildrop.html|maildrop]] * [[http://www.wonkity.com/~wblock/docs/html/maildrop.html|Using maildrop As A Mail Delivery Agent (MDA)]] * [[http://manpages.ubuntu.com/manpages/trusty/man7/maildropex.7.html|.mailfilter examples]] * [[https://www.axllent.org/docs/view/gmail-pop3-with-fetchmail/|Setting up fetchmail with Gmail]] * [[https://revadig.blogspot.com/2017/08/fetchmail-check-new-email-in-gmail.html|Fetchmail, Check new email in gmail, yahoo and other accounts]] * [[https://www.howtoforge.com/debian_etch_fetchmail|Retrieving Emails From Remote Servers With fetchmail (Debian Etch)]] * [[https://www.linode.com/docs/email/clients/using-fetchmail-to-retrieve-email/|Using Fetchmail to Retrieve Email]] * [[http://www.unixmantra.com/2013/04/aix-powerha-hacmp-commands.html|Useful PowerHA (hacmp) commands]] Eventually, after months of trying different things, came up with this combination of tools that worked very well. === Install tools === sudo dnf -y install fetchmail maildrop uudeview openssl * Set Gmail account up for POP and disable IMAP === Configure fetchmail === This is the part that talks to Gmail and downloads the emails to the local machine. vi ~/.fetchmailrc set postmaster "" #set daemon 600 poll pop.gmail.com with proto POP3 user "account@gmail.com" there with password '' is here options ssl nokeep mda "/usr/bin/maildrop .mailfilter" # sslcertfile /etc/ssl/certs/ca-bundle.crt If fetchmail complains about ssl certificates, try specifying wherein is with the sslcertfile line === Configure maildrop === maildirmake maildrop maildirmake backup vi ~/.mailfilter LOGFILE = "/home/dbahawk/maildrop/maildrop.log" DEFAULT="$HOME/maildrop" # keep a copy cc backup `cd backup/new && rm -f dummy \`ls -t | sed -e 1,50d\`` if (/^To:.*getmail@...\.dnsalias\.org$/) { to Maildir/getmail/ } #if (/^To:.*dbahawk/) # { # to maildrop # } if (/^To:.*dbahawk/) dotlock "auto.lock" { to "|uudeview -c -i" } Not really sure why this filter is not working properly. The cc works but the To: check doesn't. So I have a shell that runs separately to process the emails in the backup mail directory. === uudeview === This is a clever program that can extract attachments from emails and put them in a directory of you choosing.\\ This shell runs from cron at regular intervals to run the whole process #!/usr/bin/ksh # fetch any new mail from Gmail (uses .fetchmailrc) # fetchmail log is $HOME/maildrop/maildrop.log fetchmail -d0 # fetchmail uses maildrop as mda (uses .mailfilter) to save messages to a maildir. # this is not working correctly but still copies the messages to backup # pull out the attachments and send them to the incoming directory for the collector to process when it wakes up uudeview -c -i -p ~/Downloads/dbahawk_forward backup/new/* ==== Reset terminal to a sane state if it has got messed up ==== If, for example, you have done a cat on an binary file, sometines the terminal gets screwed up. This can get it back. stty sane The use of CTRL + J (Linefeed) is in case the Return key has also stopped working. ==== The underused xargs command! ==== See man xargs ls | xargs -t -I {} mv {} {}.old This command sequence renames all files in the current directory by adding .old to the end of each name. The -I flag tells the xargs command to insert each line of the ls directory listing where {} (braces) appear. If the current directory contains the files chap1, chap2, and chap3, this constructs the following commands: mv chap1 chap1.old mv chap2 chap2.old mv chap3 chap3.old ==== Watch a directory and automatically perform an action when new files come into it ==== Using a watchdog program like this Python script can do the trick ([[https://www.michaelcho.me/article/using-pythons-watchdog-to-monitor-changes-to-a-directory|Using Python's Watchdog to monitor changes to a directory]]) import time from watchdog.observers import Observer from watchdog.events import FileSystemEventHandler class Watcher: DIRECTORY_TO_WATCH = "/path/to/my/directory" def __init__(self): self.observer = Observer() def run(self): event_handler = Handler() self.observer.schedule(event_handler, self.DIRECTORY_TO_WATCH, recursive=True) self.observer.start() try: while True: time.sleep(5) except: self.observer.stop() print "Error" self.observer.join() class Handler(FileSystemEventHandler): @staticmethod def on_any_event(event): if event.is_directory: return None elif event.event_type == 'created': # Take any action here when a file is first created. print "Received created event - %s." % event.src_path elif event.event_type == 'modified': # Taken any action here when a file is modified. print "Received modified event - %s." % event.src_path if __name__ == '__main__': w = Watcher() w.run() ==== Top 10 (or 20) thread count consumers in AIX ==== ps -ef -o thcount,user,pid,ppid,time,args|sort -nr|head -20 ==== How much memory is being used on AIX? ==== * [[https://www.ibm.com/support/pages/aix-memory-usage-or-who-using-memory-and-how/|AIX Memory Usage - or - Who is using the memory and how?]] * [[https://www.ibm.com/support/pages/aix-memory-usage-100-used-filecache-and-paging|AIX Memory usage: 100% used, filecache and paging]] * [[https://developer.ibm.com/components/aix/articles/using-svmon-to-display-available-memory-on-aix/|Using svmon to display available memory on IBM AIX]] svmon -O summary=basic,unit=GB Unit: GB -------------------------------------------------------------------------------------- size inuse free pin virtual available mmode memory 270.00 266.92 3.08 22.9 204.16 57.2 Ded pg space 64.0 0.84 work pers clnt other pin 14.7 0 0.04 8.18 in use 204.16 0 62.8 ==== Top memory consumers in AIX ==== svmon -U -t 10 -O process=on -O sortentity=pgsp ==== Top 10 memory-consuming processes ==== ps aux | head -1; ps aux | sort -rn +3 | head or, in order of real memory usage ps vx | head -1; ps vx | grep -v PID | sort -rn +6 | head -10 ==== Show paged memory hogs on AIX ==== Show top 10 processes using memory svmon -Pt20 | perl -ne 'print if($.==2||$&&&!$s++);$.=0 if(/^-+$/)' ==== Top 10 processes in order of I/O usage ==== ps vx | head -1; ps vx | grep -v PID | sort -rn +4 | head -10 ==== Top 10 CPU consuming processes ==== Info on High(100%) used cpu myth! [[http://www.dba-oracle.com/t_high_cpu.htm|"Chicken Little" myth - dba-oracle.com]] On AIX ps aux | head -1; ps aux | sort -rn +2 | head -10 then ps aeww or on Linux top -b -n 1 | sed -e "1,6d" | head -11 gives something like... PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 29185 oracle 20 0 6597120 2.2g 74936 S 6.2 7.0 11:18.92 java 76486 oracle 20 0 129824 2168 1744 S 6.2 0.0 566:09.68 genorapw 1 root 20 0 191692 4012 2112 S 0.0 0.0 29:21.87 systemd 2 root 20 0 0 0 0 S 0.0 0.0 0:00.83 kthreadd 4 root 0 -20 0 0 0 S 0.0 0.0 0:00.00 kworker/0:0H 6 root 20 0 0 0 0 S 0.0 0.0 24:24.49 ksoftirqd/0 7 root rt 0 0 0 0 S 0.0 0.0 52:02.11 migration/0 8 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcu_bh 9 root 20 0 0 0 0 S 0.0 0.0 94:06.86 rcu_sched 10 root 0 -20 0 0 0 S 0.0 0.0 0:00.00 lru-add-drain Now we have the process 12385 showing 119% of cpu usage! oracle 12385 12327 99 Aug17 ? 74-09:39:08 /cln/exp/ora_bin1/app/oracle/middleware/oracle_common/jdk/bin/java -server -Xms256M -Xmx1740M -XX:PermSize=128M -XX:MaxPermSize=768M -XX:CompileThreshold=8000 -XX:-DoEscapeAnalysis -XX:+UseCodeCacheFlushing -XX:ReservedCodeCacheSize=100M -XX:+UseConcMarkSweepGC -XX:+UseParNewGC -XX:+CMSClassUnloadingEnabled -Dweblogic.Name=EMGC_OMS1 -Djava.security.policy=/cln/exp/ora_bin1/app/oracle/middleware/wlserver/server/lib/weblogic.policy -Dweblogic.ProductionModeEnabled=true -Dweblogic.system.BootIdentityFile=/cln/exp/ora_bin1/app/oracle/gc_inst/user_projects/domains/GCDomain/servers/EMGC_OMS1/data/nodemanager/boot.properties -Dweblogic.nodemanager.ServiceEnabled=true -Dweblogic.nmservice.RotationEnabled=true -Dweblogic.security.SSL.ignoreHostnameVerification=true -Dweblogic.ReverseDNSAllowed=false -DINSTANCE_HOME=/cln/exp/ora_bin1/app/oracle/gc_inst/em/EMGC_OMS1 -DORACLE_HOME=/cln/exp/ora_bin1/app/oracle/middleware -Ddomain.home=/cln/exp/ora_bin1/app/oracle/gc_inst/user_projects/domains/GCDomain -Djava.awt.headless=true -Ddomain.name=GCDomain -Doracle.sysman.util.logging.mode=dual_mode -Djbo.doconnectionpooling=true -Djbo.txn.disconnect_level=1 -Docm.repeater.home=/cln/exp/ora_bin1/app/oracle/middleware -Djbo.ampool.minavailablesize=1 -Djbo.ampool.timetolive=-1 -Djbo.load.components.lazily=true -Djbo.max.cursors=5 -Djbo.recyclethreshold=50 -Djbo.ampool.maxavailablesize=50 -Djavax.xml.bind.JAXBContext=com.sun.xml.bind.v2.ContextFactory -Djava.security.egd=file:///dev/./urandom -Dweblogic.debug.DebugWebAppSecurity=true -Dweb ogic.SSL.LoginTimeoutMillis=300000 -Djps.auth.debug=true -Djps.authz=ACC -Djps.combiner.optimize.lazyeval=true -Djps.combiner.optimize=true -Djps.subject.cache.key=5 -Djps.subject.cache.ttl=600000 -Doracle.apm.home=/cln/exp/ora_bin1/app/oracle/middleware/apm/ -DAPM_HELP_FILENAME=oesohwconfig.xml -Dweblogic.data.canTransferAnyFile=true -Dhttps.protocols=TLSv1,TLSv1.1,TLSv1.2 -DHTTPClient.retryNonIdempotentRequest=false -Dweblogic.security.SSL.minimumProtocolVersion=TLSv1 -Djava.endorsed.dirs=/cln/exp/ora_bin1/app/oracle/middleware/oracle_common/jdk/jre/lib/endorsed:/cln/exp/ora_bin1/app/oracle/middleware/oracle_common/modules/endorsed -Djava.protocol.handler.pkgs=oracle.mds.net.protocol -Dopss.version=12.1.3 -Digf.arisidbeans.carmlloc=/cln/exp/ora_bin1/app/oracle/gc_inst/user_projects/domains/GCDomain/config/fmwconfig/carml -Digf.arisidstack.home=/cln/exp/ora_bin1/app/oracle/gc_inst/user_projects/domains/GCDomain/config/fmwconfig/arisidprovider -Doracle.security.jps.config=/cln/exp/ora_bin1/app/oracle/gc_inst/user_projects/domains/GCDomain/config/fmwconfig/jps-config.xml -Doracle.deployed.app.dir=/cln/exp/ora_bin1/app/oracle/gc_inst/user_projects/domains/GCDomain/servers/EMGC_OMS1/tmp/_WL_user -Doracle.deployed.app.ext=/- -Dweblogic.alternateTypesDirectory=/cln/exp/ora_bin1/app/oracle/middleware/oracle_common/modules/oracle.ossoiap_12.1.3,/cln/exp/ora_bin1/app/oracle/middleware/oracle_common/modules/oracle.oamprovider_12.1.3,/cln/exp/ora_bin1/app/oracle/middleware/oracle_common/modules/oracle.jps_12.1.3 -Doracle.mds.filestore.preferred= -Dadf.version=12.1.3 -Dweblogic.jdbc.remoteEnabled=false -Dcommon.components.home=/cln/exp/ora_bin1/app/oracle/middleware/oracle_common -Djrf.version=12.1.3 -Dorg.apache.commons.logging.Log=org.apache.commons.logging.impl.Jdk14Logger -Ddomain.home=/cln/exp/ora_bin1/app/oracle/gc_inst/user_projects/domains/GCDomain -Doracle.server.config.dir=/cln/exp/ora_bin1/app/oracle/gc_inst/user_projects/domains/GCDomain/config/fmwconfig/servers/EMGC_OMS1 -Doracle.domain.config.dir=/cln/exp/ora_bin1/app/oracle/gc_inst/user_projects/domains/GCDomain/config/fmwconfig -Dohs.product.home=/cln/exp/ora_bin1/app/oracle/middleware/ohs -da -Dwls.home=/cln/exp/ora_bin1/app/oracle/middleware/wlserver/server -Dweblogic.home=/cln/exp/ora_bin1/app/oracle/middleware/wlserver/server -Djavax.management.builder.initial=weblogic.management.jmx.mbeanserver.WLSMBeanServerBuilder -Dxdo.server.config.dir=/cln/exp/ora_bin1/app/oracle/gc_inst/user_projects/domains/GCDomain/config/bipublisher -DXDO_FONT_DIR=/cln/exp/ora_bin1/app/oracle/middleware/bi/common/fonts -Dweblogic.management.server=https://hn1627.cln.be:7102 -Djava.util.logging.manager=oracle.core.ojdl.logging.ODLLogManager -Dweblogic.utils.cmm.lowertier.ServiceDisabled=true weblogic.Server Digging deeper. Process 12385 is a java program (WebLogic server) and is the OMS component of Oracle Enterprise Manager (see -Dweblogic.Name=EMGC_OMS1). We can find which thread of the process is using most cpu with... ps -eLo pid,ppid,tid,pcpu,comm |sort -k4| grep 12385 gives 12385 12327 80591 0.0 java 12385 12327 81262 0.0 java 12385 12327 81265 0.0 java 12385 12327 81297 0.0 java 12385 12327 88307 0.0 java 12385 12327 8833 0.0 java 12385 12327 90626 0.0 java 12385 12327 90627 0.0 java 12385 12327 93238 0.0 java 12385 12327 93314 0.0 java 12385 12327 94330 0.0 java 12385 12327 94454 0.0 java 12385 12327 98821 0.0 java 12385 12327 98824 0.0 java 12385 12327 12391 0.1 java 12385 12327 47759 0.2 java 12385 12327 12399 0.3 java 12385 12327 47962 0.5 java 12385 12327 47963 0.5 java 12385 12327 12388 17.3 java 12385 12327 12389 17.3 java 12385 12327 12390 17.3 java 12385 12327 12387 17.4 java 12385 12327 12392 2.2 java 12385 12327 51148 24.0 java 12385 12327 102611 40.4 java so thread id 102611 of process 12385 is using 40% cpu As this is Enterprise Manager, instead of running kill -3 12385 we run $AGENT_HOME/oracle_common/jdk/bin/jstack -F 12385 >stacktrace_12385.txt Look up the thread inside this trace file Thread 28550: (state = BLOCKED) - java.lang.Object.wait(long) @bci=0 (Compiled frame; information may be imprecise) - java.util.TimerThread.mainLoop() @bci=201, line=552 (Compiled frame) - java.util.TimerThread.run() @bci=1, line=505 (Interpreted frame) Thread 93314: (state = BLOCKED) - java.lang.Object.wait(long) @bci=0 (Compiled frame; information may be imprecise) - java.util.TimerThread.mainLoop() @bci=201, line=552 (Compiled frame) - java.util.TimerThread.run() @bci=1, line=505 (Interpreted frame) Thread 102611: (state = BLOCKED) - java.lang.Long.valueOf(long) @bci=27, line=577 (Compiled frame) - oracle.adfinternal.view.faces.activedata.LongPollingChannelHandler._longPollEvents(javax.faces.context.ExternalContext, oracle.adfinternal.view.faces.activedata.PageDataUpdateManager, javax.servlet.http.HttpServletResponse, oracle.adfinternal.view.faces.activedata.AdsProcessor$ADSRequestParams) @bci=243, line=324 (Compiled frame) - oracle.adfinternal.view.faces.activedata.LongPollingChannelHandler.flushActiveData(javax.faces.context.ExternalContext, oracle.adfinternal.view.faces.activedata.PageDataUpdateManager, javax.servlet.http.HttpServletResponse, oracle.adfinternal.view.faces.activedata.AdsProcessor$ADSRequestParams) @bci=6, line=88 (Interpreted frame) - oracle.adfinternal.view.faces.activedata.AdsProcessor.flushActiveData(javax.servlet.http.HttpServletRequest, javax.servlet.http.HttpServletResponse) @bci=72, line=126 (Interpreted frame) - oracle.adfinternal.view.faces.activedata.ActiveDataFlusher.run() @bci=28, line=26 (Interpreted frame) - weblogic.work.j2ee.J2EEWorkManager$WorkWithListener.run() @bci=72, line=184 (Interpreted frame) - weblogic.work.ExecuteThread.execute(java.lang.Runnable) @bci=34, line=311 (Compiled frame) - weblogic.work.ExecuteThread.run() @bci=42, line=263 (Compiled frame) Thread 119236: (state = BLOCKED) - java.lang.Object.wait(long) @bci=0 (Compiled frame; information may be imprecise) - java.lang.Object.wait() @bci=2, line=503 (Compiled frame) - oracle.sysman.util.threadPoolManager.Worker.getWork() @bci=8, line=132 (Compiled frame) - oracle.sysman.util.threadPoolManager.WorkerThread.run() @bci=21, line=304 (Compiled frame) Need to contact Oracle Support about this one... Update: Before it was resolved, the system was shutdown and restarted as part of a DR exercise. Problem has gone away :-) ==== How can I either encrypt or render my shell script unreadable? ==== * [[https://unix.stackexchange.com/questions/90178/how-can-i-either-encrypt-or-render-my-shell-script-unreadable]] * [[https://www.thegeekstuff.com/2012/05/encrypt-bash-shell-script/|How to Encrypt Your Bash Shell Script on Linux Using SHC]] Usage $ ./shc -f random.sh Once you run it your shell script, random.sh will get converted into this file: -rwx-wx--x. 1 ramesh ramesh 11752 Mar 27 01:12 random.sh.x ==== Date arithmetic ==== In bash, take one hour off the current time printf "current date: %(%m/%d/%Y -%H:%M:%S)T\n" printf "date - 60min: %(%m/%d/%Y -%H:%M:%S)T\n" $(( $(printf "%(%s)T") - 60 * 60 )) ==== Out of memory saving lines for undo - try using ed ==== vi fails to open (semi) large files. Put this in .kshrc export EXINIT="set ll=80000000" ==== Restart a program / service / daemon if it has not been seen for a while ==== Use the %s format of the date command to be able to to date arithmetic #!/bin/ksh MAX_DELTA=60 last_occurance_txt=$(/bin/grep sssd /var/log/messages | /bin/grep "Preauthentication failed" | /usr/bin/tail -1 | /bin/cut -c1-15) if [ ! -z "$last_occurance_txt" ] then last_occurance=$(/bin/date --date="$last_occurance_txt" +%s) now=$(/bin/date +%s) delta=$((now - last_occurance)) if [ $delta -le $MAX_DELTA ] then /bin/date /sbin/service sssd restart fi fi ==== Start / Stop NFS mounts ==== #!/bin/ksh if test "x$GROUPNAME" == "x" then echo "GROUPNAME does not exist script was not called by PowerHA" exit 16 fi echo "NFS GROUPNAME=$GROUPNAME" function _usage { echo "usage: $1 {start|stop|list}" echo " start : Mount the NFS for RG" echo " stop : Unmount the NFS for RG" } if [ $# -ne 1 ] ; then _usage "$0" exit 1 fi case "$1" in start) mount -t $GROUPNAME ;; stop) umount -f -t $GROUPNAME ;; *) _usage "$0" exit 1 ;; esac ==== Start / Stop / List Samba shares ==== #!/bin/ksh DIRECT=$(dirname "$0") HACMP_RG_NAME=$(echo $DIRECT | cut -d"/" -f4) echo $HACMP_RG_NAME | grep -q HN rc=$? if [ $rc -ne 0 ] then echo "Must be started with full path (/ux/hacmp/HN.../rc2.d/...)" exit fi HACMP_RG_NAME2=$((cd /ux/hacmp ; ls -d HN??? )| grep -vw $HACMP_RG_NAME) LNN=`/usr/es/sbin/cluster/utilities/get_local_nodename|tr "[:lower:]" "[:upper:]"` HACMP_REMOTE_HOST=$(/usr/es/sbin/cluster/utilities/cllsnode | grep "^Node" | awk '{print $2}' | grep -vw $(/usr/es/sbin/cluster/utilities/get_local_nodename)| tr "[:lower:]" "[:upper:]") SMBCONF=/ux/hacmp/${HACMP_RG_NAME}/fil/smb.conf SMBPID=/ux/hacmp/${HACMP_RG_NAME}/run/smbd-smb.conf.pid NMBPID=/ux/hacmp/${HACMP_RG_NAME}/run/nmbd-smb.conf.pid WINPID=/ux/hacmp/${HACMP_RG_NAME}/run/winbindd-smb.conf.pid IP_SERVICE=$(grep -w ^interfaces $SMBCONF | cut -d"=" -f2 | awk '{print $1}' | cut -d"/" -f1) INTERFACE=${IP_SERVICE} function _usage { echo "usage: $1 {start|stop|list}" echo " start : start the SAMBA deamon" echo " stop : start the SAMBA deamon" echo " list : list the SAMBA shares" } if [ $# -ne 1 ] ; then _usage "$0" exit 1 fi case "$1" in [sS][tT][aA][rR][tT]) # mkdev -l posix_aio0 echo 'Starting smbd ...' /opt/pware64/sbin/smbd -s "$SMBCONF" -D echo 'Starting nmbd ...' /opt/pware64/sbin/nmbd -s "$SMBCONF" -D echo 'Starting winbindd ...' /opt/pware64/sbin/winbindd -s "$SMBCONF" ;; [sS][tT][oO][pP]) # a nice kill of SAMBA processes for proc in $(cat $SMBPID $NMBPID $WINPID) do if [ -f $proc ] then echo "Sending SIGTERM to the process $proc ..." kill -TERM $proc > /dev/null 2>/dev/null sleep 1 fi done ps -ef|grep pware|grep $HACMP_RG_NAME|grep -v grep|grep -v stop if [ $? == 0 ] then for proc in `ps -ef|grep pware|grep $HACMP_RG_NAME|grep -v grep|awk '{print $2}'` do kill -9 $proc done fi # if [ $LNN == $HACMP_REMOTE_HOST ] # then # /ux/hacmp/${HACMP_RG_NAME2}/init.d/samba.ksh stop;/ux/hacmp/${HACMP_RG_NAME2}/init.d/samba.ksh start # fi ;; [lL][iI][sS][tT]) /opt/pware64/bin/smbclient -L $INTERFACE -N ;; info) echo "boot local $LNN" echo "boot remote $HACMP_REMOTE_HOST" echo "service local $HACMP_RG_NAME" echo "service remote $HACMP_RG_NAME2" echo "configuration $SMBCONF" echo "IP_SERVICE $IP_SERVICE" echo "INTERFACE $INTERFACE" ;; *) _usage "$0" exit 1 ;; esac exit 0 ==== Script called at node startup using a file on a shared cluster filesystem to start Oracle databases and listeners ==== (0) nodb_121 oracle@hn5205:/home/oracle> sudo cat /ux/hacmp/init.d/oracle.ksh #!/bin/ksh DIRECT=$(dirname "$0") # some configuration variables : ORATAB=/ux/hacmp/${HACMP_RG_NAME}/file/oratab_${HACMP_RG_NAME} LSNRDESC=/ux/hacmp/${HACMP_RG_NAME}/file/lsnrdesc_${HACMP_RG_NAME} echo "ORATAB=$ORATAB" echo "LSNRDESC=$LSNRDESC" function _usage { echo "usage: $1 {start|stop}" echo " start : start Oracle DBs" echo " stop : stop Oracle DBs" } if [ $# -ne 1 ] ; then _usage "$0" exit 1 fi case "$1" in [sS][tT][aA][rR][tT]) echo 'Starting Oracle ...' export ORATAB LSNRDESC echo 'Starting listener ...' for i in `sed -e 's/#.*$//g' -e '/^$/d' $LSNRDESC | awk -F":" '{print $1}'` do ORACLE_OWNER=`cat $LSNRDESC | awk -F":" '{if($1=="'$i'")print $4}'` export ORACLE_OWNER ORALOG=/uxlog/hacmp/${HACMP_RG_NAME}/oracle/start_${i}.log /ux/hacmp/init.d/rc.oracle.ksh start lsnr $i 1>$ORALOG 2>&1 & done wait sleep 10 echo 'Starting DB ...' for i in `sed -e 's/#.*$//g' -e '/^$/d' $ORATAB | awk -F":" '{print $1}'` do ORACLE_OWNER=`cat $ORATAB | awk -F":" '{if($1=="'$i'")print $4}'` export ORACLE_OWNER ORALOG=/uxlog/hacmp/${HACMP_RG_NAME}/oracle/start_${i}.log /ux/hacmp/init.d/rc.oracle.ksh start db $i 1>$ORALOG 2>&1 & done wait ;; [sS][tT][oO][pP]) echo 'Stopping Oracle ...' export ORATAB LSNRDESC echo 'Stopping DB ...' for i in `cat $ORATAB | awk -F":" '{print $1}'` do ORACLE_OWNER=`sed -e 's/#.*$//g' -e '/^$/d' $ORATAB | awk -F":" '{if($1=="'$i'")print $4}'` export ORACLE_OWNER ORALOG=/uxlog/hacmp/${HACMP_RG_NAME}/oracle/stop_${i}.log /ux/hacmp/init.d/rc.oracle.ksh stop db $i 1>$ORALOG 2>&1 & done wait echo 'Stopping listener ...' for i in `cat $LSNRDESC | awk -F":" '{print $1}'` do ORACLE_OWNER=`sed -e 's/#.*$//g' -e '/^$/d' $LSNRDESC | awk -F":" '{if($1=="'$i'")print $4}'` export ORACLE_OWNER ORALOG=/uxlog/hacmp/${HACMP_RG_NAME}/oracle/stop_${i}.log /ux/hacmp/init.d/rc.oracle.ksh stop lsnr $i 1>$ORALOG 2>&1 & done wait ;; *) _usage "$0" exit 1 ;; esac exit 0 ==== Script called by previous one to start databases and / or listeners (supplied by Oracle Corp. many eons ago ==== (0) nodb_121 oracle@hn5205:/home/oracle> sudo cat /ux/hacmp/init.d/rc.oracle.ksh #!/bin/ksh ######################################################################### # # # Name: rc.oracle # # # # Version: 24/11/2008 # # # # Description: This script starts and stops ORACLE V7, V8 and # # V9 databases as well as SQLNET listeners. # # It processes only the databases/listeners # # whose entry in the file $ORATAB/$LSNRDESC has # # the third field set to Y. # # The original purpose of this script was to # # start and stop all databases and all listeners # # defined in the files $ORATAB and $LSNRDESC. # # It is however possible to start and stop only # # the databases or only the listeners, only one # # database or only one listener by specifying the # # appropiated arguments. # # A database is started with the "mount" option. # # If there is any tablespace in backup mode, its # # status is reset before the database is opened. # # For each database to be stopped, a "shutdown # # immediate" is performed in background. Before # # stopping the listeners, the script waits for # # the completion of the shutdowns. After a # # timeout of $TIME_BEF_ABORT seconds, it # # aborts the databases that are still in the # # shutdown process. # # Variables ORATAB, LSNRDESC and TIME_BEF_ABORT # # have to be initialized in this script. For the # # value of $TIME_BEF_ABORT, take into account the # # number of databases to shut and their size. # # # # Arguments: [start|stop] [db|lsnr] [DBSID|LSNRNAME] # # # # Files: $ORATAB (/etc/dbdesc for example), # # $LSNRDESC (/etc/lsnrdesc for example) # # # # Scripts: # # # # Returns: 1 if file $ORATAB or file $LSNRDESC is # # inaccessible or if bad argument received. # # 0 otherwise # # # ######################################################################### date if [ -z "${ORATAB}" -o -z "${LSNRDESC}" ] then echo "ORATAB or LSNRDESC not set" exit 1 fi TIME_BEF_ABORT=360 if [ "$VERBOSE_LOGGING" = "high" ] then set -x fi # oracle processes CLEARING kill_oracle_process () { if [ "$VERBOSE_LOGGING" = "high" ] then set -x fi sleep 5 PROC_NAME="oracle$1" ps -eo pid,args | awk '{if($2=="'$PROC_NAME'")print $1}' | while read PROC_NBR do kill -9 $PROC_NBR done } # oracle processes CHECKING oracle_proc_running () { if [ "$VERBOSE_LOGGING" = "high" ] then set -x fi SID=$1 NUMBOP=$2 X=`ps -eo args | awk '$1 ~ /ora_.*_'$SID'$/' | wc -l` if [ $X -ge $NUMBOP ] then return 0 # ORACLE RUNS FOR SID = $1 else return 1 # ORACLE DOES NOT RUN FOR SID = $1 fi } case $1 in ''|'start') TASK="START";; 'stop') TASK="STOP";; *) echo "Usage $0 [start|stop] [db|lsnr] [DBSID|LSNRNAME]"; exit 1;; esac case $2 in '') TASK=$TASK"ALL";; 'db') TASK=$TASK"DB";; 'lsnr') TASK=$TASK"LSNR";; *) echo "Usage $0 [start|stop] [db|lsnr] [DBSID|LSNRNAME]"; exit 1;; esac DLNAME="^.*:" if [ "$TASK" != "STARTALL" -a "$TASK" != "STOPALL" -a "$3" != "" ] then DLNAME="^${3}:" fi PATH=/usr/bin:/etc:/usr/sbin:/usr/bin/X11; export PATH SAVE_PATH=$PATH; export SAVE_PATH LIBPATH=/usr/lib:/usr/lib/X11; export LIBPATH SAVE_LIBPATH=$LIBPATH; export SAVE_LIBPATH AIXTHREAD_SCOPE=S; export AIXTHREAD_SCOPE if [ "$TASK" = "STARTALL" -o "$TASK" = "STARTLSNR" ] then if [ ! -r $LSNRDESC ] then echo "Can't access $LSNRDESC" exit 1 fi echo "\nORACLE listeners startup\n************************" grep "^[a-z,A-Z,0-9]" $LSNRDESC | grep $DLNAME | ; awk -F":" '{if($3=="Y")print $1,$2}' | ; while read LSNR_NAME ORACLE_HOME do if [ ! -x $ORACLE_HOME/bin/lsnrctl ] then echo "Can't find or execute $ORACLE_HOME/bin/lsnrctl" else PATH=$ORACLE_HOME/bin:$SAVE_PATH if [ -d $ORACLE_HOME/lib32 ] then LIBPATH=$ORACLE_HOME/lib:$ORACLE_HOME/lib32:$SAVE_LIBPATH else LIBPATH=$ORACLE_HOME/lib:$SAVE_LIBPATH fi LD_LIBRARY_PATH=$LIBPATH TNS_ADMIN=$ORACLE_HOME/network/admin export ORACLE_HOME PATH LIBPATH LD_LIBRARY_PATH TNS_ADMIN if [ -r $ORACLE_HOME/ops/ogms_profile.sh ] then . $ORACLE_HOME/ops/ogms_profile.sh fi if [ "$LSNR_NAME" = "DBSNMP" ] then su $ORACLE_OWNER -c "lsnrctl dbsnmp_start" else su $ORACLE_OWNER -c "lsnrctl start $LSNR_NAME" fi fi done [ "$TASK" = "STARTALL" ] && sleep 10 fi if [ "$TASK" = "STARTALL" -o "$TASK" = "STARTDB" ] then if [ ! -r $ORATAB ] then echo "Can't access $ORATAB" exit 1 fi echo "\nORACLE databases startup\n************************" grep "^[a-z,A-Z,0-9]" $ORATAB | grep $DLNAME | ; awk -F":" '{if($3=="Y")print $1,$2}' | ; while read ORACLE_SID ORACLE_HOME do PATH=$ORACLE_HOME/bin:$SAVE_PATH if [ -d $ORACLE_HOME/lib32 ] then LIBPATH=$ORACLE_HOME/lib:$ORACLE_HOME/lib32:$SAVE_LIBPATH else LIBPATH=$ORACLE_HOME/lib:$SAVE_LIBPATH fi LD_LIBRARY_PATH=$LIBPATH export ORACLE_HOME ORACLE_SID PATH LIBPATH LD_LIBRARY_PATH if [ -r $ORACLE_HOME/ops/ogms_profile.sh ] then . $ORACLE_HOME/ops/ogms_profile.sh fi export SQLDBA="" if [ -x $ORACLE_HOME/bin/svrmgrl ] then SQLDBA=svrmgrl elif [ -x $ORACLE_HOME/bin/sqlplus ] then SQLDBA="sqlplus /nolog" fi if [ "$SQLDBA" = "" ] then echo "Can't find or execute svrmgrl|sqlplus under $ORACLE_HOME/bin" elif [ ! -f $ORACLE_HOME/dbs/init${ORACLE_SID}.ora -a ! -f $ORACLE_HOME/dbs/spfile${ORACLE_SID}.ora ] then echo "Can't find init file for Database $ORACLE_SID" elif oracle_proc_running $ORACLE_SID 4 then echo "ORACLE IS ALREADY RUNNING FOR SID : $ORACLE_SID" else echo "Clearing memory for database $ORACLE_SID" su $ORACLE_OWNER -c '${SQLDBA} < /tmp/endbkp${ORACLE_SID}.sql if [ -s /tmp/endbkp${ORACLE_SID}.sql ] then echo "Resetting the status of datafile(s) found in backup mode" chmod 644 /tmp/endbkp${ORACLE_SID}.sql su $ORACLE_OWNER -c '${SQLDBA} </tmp/shut${ORACLE_SID}.sh chmod 755 /tmp/shut${ORACLE_SID}.sh su $ORACLE_OWNER -c "/tmp/shut${ORACLE_SID}.sh >/tmp/shut${ORACLE_SID}.log 2>&1 &" DBRUN=${DBRUN}${ORACLE_SID}" " fi done # Wait for completion of the shutdowns while [ "$DBRUN" != "" -a $TIME_BEF_ABORT -gt 0 ] do sleep 15 TIME_BEF_ABORT=`expr $TIME_BEF_ABORT - 15` DBRUNBIS="" for ORACLE_SID in $DBRUN do ORACLE_HOME=`grep "^${ORACLE_SID}:" $ORATAB | ; awk 'BEGIN{FS=":"}{print $2}'` if [ -f /tmp/shut${ORACLE_SID}.flag ] then kill_oracle_process ${ORACLE_SID} & echo "Oracle database $ORACLE_SID has been shut properly" else DBRUNBIS=${DBRUNBIS}${ORACLE_SID}" " fi done DBRUN=$DBRUNBIS done # Abort any database still running for ORACLE_SID in $DBRUN do ORACLE_HOME=`grep "^${ORACLE_SID}:" $ORATAB | ; awk 'BEGIN{FS=":"}{print $2}'` PATH=$ORACLE_HOME/bin:$SAVE_PATH LIBPATH=$ORACLE_HOME/lib:$SAVE_LIBPATH export ORACLE_HOME ORACLE_SID PATH LIBPATH if [ -r $ORACLE_HOME/ops/ogms_profile.sh ] then . $ORACLE_HOME/ops/ogms_profile.sh fi export SQLDBA="" if [ -x $ORACLE_HOME/bin/svrmgrl ] then SQLDBA=svrmgrl else SQLDBA="sqlplus /nolog" fi if oracle_proc_running $ORACLE_SID 1 then echo "Aborting Oracle database $ORACLE_SID" su $ORACLE_OWNER -c '${SQLDBA} < ==== Use sed to repeat characters ==== echo an '=' sign then use the loop feature of sed to run round a loop a further 79 times replacing what you've got with the same thing plus and extra one! echo "=" | sed -e :a -e 's/^=\{1,79\}$/&=/;ta' ==== Convert between lower and upper case using ansi standard ==== This way should be portable and give predictable results LOWER_SID=$(echo "${SID}" | tr '[[:upper:]]' '[[:lower:]]') UPPER_SID=$(echo "${SID}" | tr '[[:lower:]]' '[[:upper:]]') ==== Use sed to delete the first line of a file or command output printing from the second line onwards ==== Actually deceptively simple. Example remove the header from df output. df -k | sed 1d or more formally df -k | sed -e '1,1d' As a useful extension, suppose TO_DATA_DIR contains multiple directories, this will total the free space for all relevant filesystems\\ This one is specific to AIX, for other Unixes and Linux, use $4 instead of $3. SPACE_AVAILABLE=$(df -k $TO_DATA_DIR | sed 1d | awk 'BEGIN {i=0} {i=i+$3} END {print i}') ==== Use sed to munge a controlfile into submission! ==== * delete comment lines * delete all lines after the one starting with a semicolon * delete all blank lines * change reuse to set * change archivelog (with word boundaries) to noarchivelog * change old sid to new sid sqlplus -s / as sysdba<<'EOSQL' >/dev/null alter database backup controlfile to trace as '/tmp/regenerate_controlfile.sql' reuse resetlogs; EOSQL OLDSID='PROD' NEWSID='TEST' sed -e '/^--/d' -e '/^\;/q' -e '/^ *$/d' -e 's/REUSE/SET/' -e 's/\\/NOARCHIVELOG/' -e 's/\\"'${OLDSID}'\\"/\\"'${NEWSID}'\\"/' /tmp/regenerate_controlfile.sql ==== Find all files with lines containing a certain text ==== I am using this to search through all files that may contain a hard-coded password. On AIX, the -H switch to grep does not print out the filename. Using this technique, adding /dev/null to the search list, forces grep to print out the filename because it is now searching multiple files. find / -name "*sh" -exec grep -n 'password_text_to_find' /dev/null {} + 2>/dev/null ==== Find all files where a certain text does not exist ==== Mostly we want to search for files containing a particular string but how to find those files that do not contain some text.\\ Using xargs instead of -exec forces the filename to be printed. -H should do this but doesn't seem to in this case. find . -name "*cfg" | xargs grep -H -c TO_CV_HOST {} \; | grep ':0$' | cut -d: -f1 ==== tar and gzip in one command using AIX ==== tar on AIX does not have the -z flag so it has to be done manually cd $ORACLE_HOME/../.. tar -cvf - 19.12/ | gzip > /oracle/Patches/AIX.PPC64_19.12_gold.gz and to unzip cd $ORACLE_HOME/../.. gunzip < /oracle/Patches/AIX.PPC64_19.12_gold.gz | tar -xvf - ==== Use tar and gzip to copy files more efficiently across network between hosts ==== From the destination server, this will connect to the source, tar up each file and pull it back to the current server ssh oracle@hn5211 "cd /oracle/ora_data4/iten3/ && tar -cf - . | gzip " | ( cd /oracle/ora_data2/iten/ && gunzip -c | tar -xvf - . ) ==== double hash and double percent in shell variables to trim off characters from variables ==== * # - trims from the start until the first occurrence of those characters matched by what follows the hash (abstemious) :-) * ## - trims from the start until the last occurrence of those characters matched by what follows the hash (greedy) * % - trims from the end backwards until the first occurrence of those characters matched by what follows the percent (abstemious) * %% - trims from the end backwards until the last occurrence of those characters matched by what follows the percent (greedy) FILENAME="/home/bey9at77/hello.txt" FILE_STUB1=${FILENAME##*\/} FILE_STUB=${FILE_STUB1%\.*} echo $FILE_STUB hello FILE_EXT=${FILENAME##*\.} echo $FILE_EXT txt or\\ # - remove prefix reluctantly\\ ## - remove prefix greedily\\ % - remove suffix reluctantly\\ %% - remove suffix greedily\\ words="do.re.mi" echo ${words#*\.} re.mi echo ${words##*\.} mi echo ${words%\.*} do.re echo ${words%%\.*} do ==== The best shell script ever. An example of how scripting should be done ==== It was written by someone at Oracle. Unfortunately (s)he did not put any author comment in it. Also unfortunately I cannot show it here as it is protected behind Oracles support website.\\ If you have an Oracle Metalink id, you can get the complete script [[https://support.oracle.com/epmos/faces/DocumentDisplay?id=949322.1|here]]\\ Here is a snippet of a function that demonstrates proper commenting and a very good style.
The script is called physru.sh and upgrades an Oracle database in a rolling upgrade fashion by using a physical standby.\\ There are 4500 lines in the full script but it is so easy to read and understand because of the way it's written, it's like a breath of fresh air. Well done whoever you are! ########################################################################### # NAME: wait_mrp_active # # DESCRIPTION: # Wait for $MRP_START_TIMEOUT minutes to confirm that the MRP is active. If # we can't detect an active MRP, abort the script with an error. # # INPUT(S): # Arguments: # $1: database user # $2: user password # $3: tns service name # $4: database unique name # # Globals: # None # # RETURN: # None # ########################################################################### wait_mrp_active() { display "confirming media recovery is running" l_wma_status=1 l_wma_curtime=`perl -e 'print int(time)'` l_wma_exptime=`expr $MRP_START_TIMEOUT "*" 60` l_wma_maxtime=`expr $l_wma_curtime + $l_wma_exptime` while [[ "$l_wma_curtime" -lt "$l_wma_maxtime" ]] do is_mrp_running $1 $2 $3 $4 if [[ "$?" -gt "0" ]]; then l_wma_status=0 break fi sleep $MRP_START_INTERVAL l_wma_curtime=`perl -e 'print int(time)'` done chkerr $l_wma_status "could not detect an active MRP after $MRP_START_TIMEOUT minutes" } ==== Use Input Field Separator (IFS) to split a set of numbers into its constituent parts ==== This example splits the MacOS version into major and minor numbers OLDIFS=$IFS IFS='.' read osvers_major osvers_minor osvers_dot_version <<< "$(/usr/bin/sw_vers -productVersion)" IFS=$OLDIFS ==== Split a file into pieces ==== This will split a file into separate files with 20000 lines in each (except maybe the last one!) awk 'NR%20000==1{x="alert_lbk5t.log."++i;}{print > x}' alert_lbk5t.log ==== Strip out duplicate lines with this clever awk trick ==== Assigns each line to an associative array element and prints the line if it does not already exist in the array awk '!visited[$0]++' your_file > deduplicated_file ==== split string (eg. filename) into separate variables using set ==== FILENAME="split_this_into_bits.txt" set $(echo "${FILENAME}" | sed 's/_/ /g') echo $4 $3 $2 $1 $5 bits.txt into this split ==== Centre align text on a line in bash shell ==== Doesn't work in Korn shell due to %*s #COLUMNS=$(tput cols) # width of the current window COLUMNS=80 title="Hello world!" printf "%*s\n" $(((${#title}+$COLUMNS)/2)) "$title" and as a one-liner printf "%*s\n" $(( ( $(echo $* | wc -c ) + 80 ) / 2 )) "$*" This works in Korn shell... TITLE="$1" LINEWIDTH=80 LPAD=$(((${#TITLE}+$LINEWIDTH)/2)) printf %${LPAD}s "$TITLE" ==== Right justify text (or repeat characters) on a line with leader dots! ==== function rpadwait { text=$1 # ------------------------- # suppress newline for echo # ------------------------- N= C= if echo "\c" | grep c >/dev/null 2>&1; then N='-n' else C='\c' fi echo ${N} "${text}${C}" | sed -e :a -e 's/^.\{1,80\}$/&\./;ta' } ==== cron jobs not submitted - daemon run out of child processes? ==== Sometimes have problems with NFS mounts and this causes cron jobs to hang. If they are scheduled to run regularly, eventually cron will no longer be able to start any more jobs.\\ * Check the cron log /var/adm/cron/log to see if there are any errors or other messages around the time the jobs should run. If cron has hit its process limit (default 100), it will try again after a number of seconds (default 60).\\ Both the number of jobs and wait time are configured in the file /var/adm/cron/queuedefs. If it is unusual for cron to be running so many jobs, you can check the process table to view the jobs cron has created. These jobs will have parent process id (PPID) of the cron daemon.\\ $ ps -ef | grep cron | grep -v grep root 6750314 1 0 Apr 24 # 3:39 /usr/sbin/cron solax025:root[/home/root]# ps -T 6750314 PID TTY TIME CMD 6750314 # 3:39 cron 21168296 # 0:00 \--bsh 58982414 # 0:00 \--sadc In this example, we only have 1 job running ==== Find long-running processes with a cron job ==== Processes running longer than 24 hours have a date instead of a start time... 58 08,14 * * * /home/ibmtools/scripts/oracle/dosh -vc "ps -ef|grep 'ibmtools/scripts/oracle'|perl -nae 'print if \$F[4] !~/:/'" >/tmp/lrp.txt; [[ $(grep -c ibmtools /tmp/lrp.txt) -ne 0 ]] && cat /tmp/lrp.txt|mailx -s '*** long running processes - please check ***' reporting@companymail.com ==== Process command line arguments in shell ==== Borrowed from Mozilla Firefox installer # Command line arg defaults # moz_debug=0 moz_debugger="" moz_debugger_args="" # # # Parse the command line # while [[ $# -gt 0 ]] do case $1 in -g | --debug) moz_debug=1 shift ;; -d | --debugger) moz_debugger=$2; if [[ "${moz_debugger}" != "" ]]; then shift 2 else echo "-d requires an argument" exit 1 fi ;; -a | --debugger-args) moz_debugger_args=$2; if [[ "${moz_debugger_args}" != "" ]]; then shift 2 else echo "-a requires an argument" exit 1 fi ;; *) break; ;; esac done # # # Program name given in $1 # if [[ $# -gt 0 ]]; then MOZ_PROGRAM=$1 shift fi # # Program not given, try to guess a default # if [[ -z "$MOZ_PROGRAM" ]]; then ## ## Try this script's name with '-bin' appended ## if [[ -x "$MOZ_DEFAULT_NAME" ]]; then MOZ_PROGRAM=$MOZ_DEFAULT_NAME ## ## Try mozilla-bin ## elif [[ -x "$MOZ_APPRUNNER_NAME" ]]; then MOZ_PROGRAM=$MOZ_APPRUNNER_NAME fi fi # # # # Make sure the program is executable # if [[ ! -x "$MOZ_PROGRAM" ]]; then moz_bail "Cannot execute $MOZ_PROGRAM." fi ==== Carry XWindows settings across sessions ==== # ---------------------- # push XWindows settings # ---------------------- [[ "$(uname)" == "SunOS" ]] && PATH=/usr/openwin/bin:$PATH WHOAMI=$(id | awk -F')' '{print $1}' | awk -F'(' '{print $2}') xauth list > /tmp/xauth_list_$WHOAMI chmod 777 /tmp/xauth_list_$WHOAMI echo $DISPLAY > /tmp/xdisplay_$WHOAMI chmod 777 /tmp/xdisplay_$WHOAMI ==== Cross-platform version of whoami ==== WHOAMI=$(id | awk -F')' '{print $1}' | awk -F'(' '{print $2}') ==== Set terminal title from command line ==== Put something like this in the .profile\\ -n do not output the trailing newline\\ -e enable interpretation of backslash escapes\\ |0 sets title of window and icon\\ |1 sets title of icon only\\ |2 sets title of window only\\ echo -en "\033]0;`hostname`\007" echo -en "\033]2;`hostname`\007" ==== Remove blank lines and comments (also indented ones) from a file ==== awk -F\: '!/^($|[:space:]*#)/ {print $2}' /etc/oratab | sort | uniq or as part of a script that removes comments and blank lines from all Korn shell scripts in a directory #!/usr/bin/ksh for i in *ksh; do perl -p -i -e 's/^\s*#[^!]*$//; s/^\s*$//' $i done ==== Return elements of an array in Korn shell ==== From [[http://unix.stackexchange.com/questions/188202/processing-output-from-an-sqlite-db-into-a-ksh-array-with-spaces|here]]\\ Could be used to separate the columns of an SQL select when returning to the shell\\ This approach eliminates the need to put quotes around text with spaces in it. echo $KSH_VERSION x="Red,Yellow is a color,Blue" oIFS=$IFS IFS=, y=($x) IFS=$oIFS echo ${y[1]} ==== A decent Unix Prompt ==== export PS1="`uname -n`:`whoami`[\${PWD}]# " or export PS1='($?) $'ORACLE_SID" "`whoami`"@"`uname -n`":"'$'PWD"> " export EDITOR=vi ==== Simple arithmetic ==== pipe the calculation into the shell calculator space_in_kb=$(echo $1 / 1024 | bc) Calculate the remainder (modulo) of a division calculation if [[ $(echo "${NUMBER} % 2" | bc) -eq 0 ]]; then echo "${NUMBER} is even" else echo "${NUMBER} is odd" fi or do it in awk if scientific notation maybe involved function calc { awk "BEGIN{print $*}"; } if [[ $(calc "${SPACE_USED} + ${SPACE_AVAILABLE} - ${DATABASE_SIZE") -le 0 ]]; then echo "NOK" fi ==== Script encryption and passphrase protection ==== Encrypt a shell script with the ability to execute the encrypted version * from [[http://www.commandlinefu.com/commands/browse|commandlinefu.com]] scrypt(){ [ -n "$1" ]&&{ echo '. <(echo "$(tail -n+2 $0|base64 -d|mcrypt -dq)"); exit;'>$1.scrypt;cat $1|mcrypt|base64 >>$1.scrypt;chmod +x $1.scrypt;};} ==== Virtual host configuration in Apache http.conf ==== ServerName dbamon DocumentRoot "/Volumes/data/Sites/dbamon_push" Options Includes FollowSymLinks AllowOverride All Order allow,deny Allow from all and in /etc/hosts, add... 127.0.0.1 dbamon ==== Mount a website (or any other remote resource) locally using WebDav ==== Redhat/CentOS yum install fuse-davfs2 or yum install wdfs.x86_64 Debian apt-get install davfs2 then... sudo mkdir /mnt/webdav # or whatever you'd like to call the directory sudo mount.davfs [-o option[,option]...] device mount_point In man's terms, that last line would translate to: id #suppose this returns uid=501 and gid=502 sudo mount.davfs -o 501,502 https://your/web/site /mnt/webdav Mac OSX osascript -e ' mount volume "http://username:password@webdav.address:portnum" ' or osascript -e ' mount volume "http://username@webdav.address:portnum" ' or osascript -e ' try mount volume "http://webdav.address:portnum" ' or mount -t webdav http://webdav.address:portnum /mnt/webdav # this one won't show up in the Finder Sidebar. ==== Using Borg to backup to a remote server via sshfs ==== * [[https://borgbackup.readthedocs.io/en/1.1-maint/index.html|BorgBackup is a deduplicating backup program. Optionally supporting compression and authenticated encryption]] ==== Use sshfs to mount a remote filesystem locally ==== * [https://vimeo.com/54505525?cjevent=331dfbdafd4f11e880e4005e0a180514|The Black Magic Of SSH / SSH Can Do That?]] * [[https://www.hiroom2.com/2017/08/04/fedora-26-sshfs-en/|Fedora 26: Install sshfs for SSH client]] Presuming Fedora 26+. Install sshfs sudo dnf -y install sshfs * Mount the filesystem locally mkdir borg sshfs stuart@192.168.1.11:borg borg read: Connection reset by peer * Debug the reason it does not mount the filesystem [dbahawk@fedora ~]$ sshfs -odebug,sshfs_debug,loglevel=debug -o Compression=no -o allow_root -o transform_symlinks stuart@192.168.1.11:borg borg SSHFS version 2.10 FUSE library version: 2.9.7 nullpath_ok: 0 nopath: 0 utime_omit_ok: 0 executing <-x> <-a> <-oClearAllForwardings=yes> <-ologlevel=debug> <-oCompression=no> <-2> <-s> read: Connection reset by peer [dbahawk@fedora ~]$ sshfs -odebug,sshfs_debug,loglevel=debug -o Compression=no -o allow_root -o transform_symlinks stuart@192.168.1.11:borg borg SSHFS version 2.10 FUSE library version: 2.9.7 nullpath_ok: 0 nopath: 0 utime_omit_ok: 0 executing <-x> <-a> <-oClearAllForwardings=yes> <-ologlevel=debug> <-oCompression=no> <-2> <-s> debug1: Reading configuration data /etc/ssh/ssh_config debug1: Reading configuration data /etc/ssh/ssh_config.d/05-redhat.conf debug1: Reading configuration data /etc/crypto-policies/back-ends/openssh.config debug1: /etc/ssh/ssh_config.d/05-redhat.conf line 8: Applying options for * debug1: Connecting to 192.168.1.11 [192.168.1.11] port 22. debug1: Connection established. debug1: key_load_public: No such file or directory debug1: identity file /home/dbahawk/.ssh/id_rsa type -1 debug1: key_load_public: No such file or directory debug1: identity file /home/dbahawk/.ssh/id_rsa-cert type -1 debug1: key_load_public: No such file or directory debug1: identity file /home/dbahawk/.ssh/id_dsa type -1 debug1: key_load_public: No such file or directory debug1: identity file /home/dbahawk/.ssh/id_dsa-cert type -1 debug1: key_load_public: No such file or directory debug1: identity file /home/dbahawk/.ssh/id_ecdsa type -1 debug1: key_load_public: No such file or directory debug1: identity file /home/dbahawk/.ssh/id_ecdsa-cert type -1 debug1: key_load_public: No such file or directory debug1: identity file /home/dbahawk/.ssh/id_ed25519 type -1 debug1: key_load_public: No such file or directory debug1: identity file /home/dbahawk/.ssh/id_ed25519-cert type -1 debug1: key_load_public: No such file or directory debug1: identity file /home/dbahawk/.ssh/id_xmss type -1 debug1: key_load_public: No such file or directory debug1: identity file /home/dbahawk/.ssh/id_xmss-cert type -1 debug1: Local version string SSH-2.0-OpenSSH_7.7 debug1: Remote protocol version 2.0, remote software version OpenSSH_7.4 debug1: match: OpenSSH_7.4 pat OpenSSH* compat 0x04000000 debug1: Authenticating to 192.168.1.11:22 as 'stuart' debug1: SSH2_MSG_KEXINIT sent debug1: SSH2_MSG_KEXINIT received debug1: kex: algorithm: curve25519-sha256@libssh.org debug1: kex: host key algorithm: ecdsa-sha2-nistp256 debug1: kex: server->client cipher: aes256-gcm@openssh.com MAC: compression: none debug1: kex: client->server cipher: aes256-gcm@openssh.com MAC: compression: none debug1: kex: curve25519-sha256@libssh.org need=32 dh_need=32 debug1: kex: curve25519-sha256@libssh.org need=32 dh_need=32 debug1: expecting SSH2_MSG_KEX_ECDH_REPLY debug1: Server host key: ecdsa-sha2-nistp256 SHA256:+h02thqjO0/SoVPccBz1RTudBCJ+V7g6MKu4E4CCToQ debug1: Host '192.168.1.11' is known and matches the ECDSA host key. debug1: Found key in /home/dbahawk/.ssh/known_hosts:1 debug1: rekey after 4294967296 blocks debug1: SSH2_MSG_NEWKEYS sent debug1: expecting SSH2_MSG_NEWKEYS debug1: SSH2_MSG_NEWKEYS received debug1: rekey after 4294967296 blocks debug1: SSH2_MSG_EXT_INFO received debug1: kex_input_ext_info: server-sig-algs= debug1: SSH2_MSG_SERVICE_ACCEPT received debug1: Authentications that can continue: publickey,password debug1: Next authentication method: publickey debug1: Trying private key: /home/dbahawk/.ssh/id_rsa debug1: Trying private key: /home/dbahawk/.ssh/id_dsa debug1: Trying private key: /home/dbahawk/.ssh/id_ecdsa debug1: Trying private key: /home/dbahawk/.ssh/id_ed25519 debug1: Trying private key: /home/dbahawk/.ssh/id_xmss debug1: Next authentication method: password stuart@192.168.1.11's password: debug1: Authentication succeeded (password). Authenticated to 192.168.1.11 ([192.168.1.11]:22). debug1: channel 0: new [client-session] debug1: Requesting no-more-sessions@openssh.com debug1: Entering interactive session. debug1: pledge: network debug1: client_input_global_request: rtype hostkeys-00@openssh.com want_reply 0 debug1: Sending environment. debug1: Sending env LANG = en_GB.UTF-8 debug1: Sending subsystem: sftp subsystem request failed on channel 0 read: Connection reset by peer The last 3 lines tell us that sftp server is probably not enabled on the remote host. Verify by trying sftp manually... [dbahawk@fedora ~]$ sftp stuart@192.168.1.11 stuart@192.168.1.11's password: subsystem request failed on channel 0 Connection closed [dbahawk@fedora ~]$ sftp -P10022 stuart@192.168.1.11 stuart@192.168.1.11's password: Connected to stuart@192.168.1.11. sftp> exit Aah, I forgot it was installed but listening on a different port. $ sshfs -o port=10022 stuart@192.168.1.11:borg borg $ Success. ==== Add current hostname to list of hosts on an xcat server ==== CMD="nodels" HOST=`hostname` (echo "$HOST"; $CMD) | while read server do echo "server:$server" done ==== What is todays 'Day Of the Year' number? ==== DOY=`perl -e 'print sub{$_[7]}->(localtime)+1;'` ==== Convert Julian day numbers to dates ==== for day in 8 33 36 61 64 91 96 121 126 152 155 182 187 215 218 244 247 274 279 306 309 335 338 365; do date -d "`date +%Y`-01-01 +$(( ${day} - 1 ))days" +%d-%m-%Y; done ==== Send crontab job output to a date formatted log file ==== This will run a job every 5 minutes and send the output to a file ending with a time in hours and minutes.\\ The thing to note is the escaped percent signs. This is because a % sign is interpreted by cron to mean a newline character. Everything after the first % is treated as input to the program. */5 * * * * /var/www/cgi-bin/dbamon_collector.ksh >/tmp/dbamon_collector.log.$(date "+\%H\%M") 2>&1 ==== Edit crontab file without crontab -e ==== It can happen that you need to add or modify a line in the crontab of many users or across many servers at once.\\ In principle, there's nothing wrong with modifying the crontab file directly. You just lose the advantages of file locking (and syntax checking) that crontab -e offers.\\ Here we take a backup of the current crontab, print it out, echo an extra command and ask cron to use these as input (thus overwriting the existing crontab file). Just don't run this close to midnight :-) crontab -l > /tmp/crontab.`date '+%Y%m%d'` ( cat /tmp/crontab.`date +'%Y%m%d'` echo "02 10 * * * /home/ibmtools/scripts/oracle/export_parfile.ksh -s SID -f JDBEOP1.parfile" ) | crontab - or crontab -l > /tmp/crontab.backup crontab -l > /tmp/crontab.$$ perl -p -i -e 's!backup_send_tsm_dump!backup_export2tsm!g' /tmp/crontab.$$ crontab /tmp/crontab.$$ rm /tmp/crontab.$$ or crontab -l >$HOME/crontab.$(date '+%Y%m%d') crontab -l | perl -p -e 's|/nas/software/oracle/scripts|/oracle/scripts|' | crontab ==== Use shell to convert a number in scientific notation to normal ==== var2convert='1.2345678e3' printf -v var2convert "%.f" $var2convert echo $var2convert # magic! ==== Check for jobs running longer that 24 hours ==== Run from the management server across all Unix servers. Checks the 5th column in a ps listing. If it doesn't find a : (time separator), the process is running longer than 24 hours. # --------------------------------------- # report on long running oraibm processes # --------------------------------------- # --------------------------------------- # report on long running oraibm processes # --------------------------------------- 01 17,14 * * * rm -f /tmp/lrp.txt;/home/ibmtools/scripts/oracle/dosh -vc \"ps -ef|egrep 'oraibm|scripts/oracle'>/tmp/lrp.txt;perl -nae 'print if \$F[4] !~/:/' /tmp/lrp.txt\" >>/tmp/lrp.txt;[[ $(egrep -c 'oraibm|scripts/or acle' /tmp/lrp.txt) -ne 0 ]] && cat /tmp/lrp.txt|mailx -s '*** long running processes - please check ***' bey9at77@mail.com ==== Global search and replace with perl (restricted to list of files provided by Unix find command) ==== find . -type f -exec perl -i -pe 's/something/else/g' {} \; ==== perl function to make filenames lower case ==== function lower { perl -e 'for (@ARGV) { rename $_, lc($_) unless -e lc($_); }' * } ==== From the management server, search the TNS listener port for each database on a server and make an inline substitution in the ITM config files! ==== for i in `/home/ibmtools/scripts/oracle/dosh -c "ls -al /opt/IBM/ITM/config/*rz*cfg|grep -v lrwx"|awk '{print $NF}'`; do server=`echo $i|cut -d_ -f1 | awk -F'/' '{print $NF}'` db=`echo $i|cut -d'.' -f1 | awk -F'_' '{print $NF}'` OH=`ssh $server grep "^$db" /etc/oratab|cut -d: -f2` LISTENERPORT=`ssh $server cat $OH/network/admin/listener.ora|perl -00 -ne 'print $1 if /'$db'.*PORT.*=.*(\d{4})/s'` ssh $server perl -p -i.bak -e 's/1521/'$LISTENERPORT'/' $i ssh $server ls -al ${i}* done ==== Run a job from cron every Nth day of the month ==== Example. Execute a job every third Saturday of the month.\\ Paste this into a file called calenday and put it in /usr/local/bin so it's (probably) on the PATH #!/usr/bin/ksh # ix is the week number of the month ("2"nd Friday of the month, "3"rd Tuesday of the month) # dy is the day number in Unix format (0 for Sunday, 1 for Monday, ... 6 for Saturday) # eg. "calenday 3 6" returns the date of 3rd Saturday of the month. ix=$1 dy=$2 SCHAR=$((($dy*2)+$dy+1)) ECHAR=$(($SCHAR+1)) cal `date '+%m %Y'` | egrep "\<[0-9]{1,2}\>" | cut -c${SCHAR}-${ECHAR} | xargs | awk {'print $'$ix'}' Now in crontab, you should be able to do something like this:\\ 15 20 * * * [[ `calenday 3 6` -eq `/bin/date '+%d'` ]] && su - oracle -c "run_my_backup.ksh" This will also work on some Unices..\\ We send backups to a special TSM node on the second Friday of each month. This report must run a day later - but that is not guaranteed to be the second Saturday or even the 3rd. So... 30 12 8-14 * 5 sleep 86400 && su - oracle -c "/usr/bin/perl -ne 'print if /ORX_M_SOL/ .. /^STOP/' /home/oracle/incoming/dbamon_spool_tsm_*.SOL | grep Archive | grep -v Client | mailx -s 'Monthly TSM backups' orareport@xxxxxx.com" ==== ps listing does not show start time after 24 hours ==== But you can see elapsed time using your own ps command /usr/bin/ps -eo etime,user,pid,ppid,cpu,start,tty,time,args|tail -n +2|sort ==== ps -ef cuts off args cmd column on Solaris ==== === To see more than 80 characters of the last column on Solaris === This shows all the individual arguments to the command pargs This shows the ps listing in 'compatibility' mode (there are more compatibility commands in /usr/xpg4/bin) /usr/ucb/ps auww === To see more than 80 characters of the last column on AIX === This shows the full argument listing of a process (NOTE: no minus sign!) ps eww ==== Remove blank / empty lines from vi ==== Maybe you cut and pasted a file from Windows and it's full of blank lines and Control-M's now\\ There are several methods but I think this is the easiest to remember :g/^$/d ==== Right pad a variable ==== a function like rpad in SQL but for Shell\\ function rpad { text=$1 padwidth=$2 padchar=$3 echo "$text" | sed -e :a -e 's/^.\{1,'$padwidth'\}$/&\'$padchar'/;ta' } ==== Connect to a Windows server from Linux using rdesktop ==== My remmina stopped working so rolled my own. Very simple really. Put this is a shell. tsocks rdesktop -z -P -x m -a 16 -d MAIND -u sbarkley -p ****** -r disk:SOL=$HOME/Documents/SOL -g 95% 150.251.112.38 & where... -z #enables compression -P #enables bitmap caching (saves network traffic) -x m#disables eye-candy features -a 16 - reduce colour pallete to 16 colours -d #domain to connect to -u #username -p #password -r #setup a shared folder -g #geometry (use W x H or percentage) Slight problems with rdesktop not working 100% of the time. Now using xfreerdp. Seems better... xfreerdp -g 90% --ignore-certificate --gdi sw -K -d wdcrhbp05 -u oraibm -p "`cat $HOME/scripts/.oraibm.password`" -T "wdcrhbp05_oraibm" --plugin cliprdr --plugin rdpdr --data disk:SOL:/home/bey9at77/Documents/SOL -- --plugin rdpsnd --data alsa -- 150.251.112.25 & ==== Reset your password bypassing password rules ==== must be done as root echo "user:new_password" | chpasswd ==== Sum the sizes of all files of an ls listing ==== It'll check to see if the sum of filesizes corresponds with the df -g (or h) listing (can get messed up due to open but deleted files) cd /oracle/export df -g . find . -name "*dmp*" -ls | awk '{ SUM += $7} END { print SUM/1024/1024/1024 }' ==== Mount an iso image under Linux ==== mkdir -p /mnt/cdrom mount -o loop /dev/cdrom /mnt/cdrom ==== How many processors on the machine? ==== * AIX lsdev -C|grep Process|wc -l * Solaris psrinfo -v|grep "Status of processor"|wc -l * Linux cat /proc/cpuinfo|grep processor|wc -l ==== Quick, simple, understandable example of how to use RRDTool ==== * [[RRDTool]] ==== Use expect to respond automatically to interactive programs ==== #!/usr/bin/expect -f spawn /usr/tivoli/tsm/client/oracle/bin64/tdpoconf passw -tdpo_optfile=/oracle/[lindex $argv 0]/admin/tdpo.opt expect "assword:" {send "password\r"} expect "assword:" {send "password\r"} expect "assword:" {send "password\r"} ==== Use expect to allow file copy with scp (if ssh keys are not an option) ==== #!/usr/bin/expect -f spawn scp /home/oracle/.profile oracle@hn512:/tmp/prfl set pass "thepassword" expect { password: {send "$pass\r" ; exp_continue} eof exit } ==== Send logging of a shell script to file from inside the script ==== Spotted this seemingly simple method in $FND_TOP/bin/batchmgr.sh # Save output nohup sh << end_logging >> $logfile 2>&1 & ... script contents ... end_logging ==== Log output of a shell script to both terminal and a file using a named pipes ==== or "How to send shell output to screen/stdout as well as to a logfile using tee and redirection with exec" * examples [[http://www.unix.com/shell-programming-and-scripting/85584-redirect-within-ksh.html|here on unix.com]] * general info on redirections [[http://www.catonmat.net/blog/bash-one-liners-explained-part-three/|here at catonmat.net]] Starting with this example found in Oracle script $ORACLE_HOME/root.sh # # Silent variable is set based on : # if OUI_SILENT is true or if SILENT_F is 1 # if [ "${OUI_SILENT}" = "true" -o "$SILENT_F" ] then SILENT=1 else SILENT=0 fi if [ "${OUI_STDOUT}" = "true" -o "$STDOUT_F" ] then STDOUT=1 else STDOUT=0 fi if [ $SILENT -eq 1 -a $STDOUT -eq 0 ] then $ECHO "Check $LOG for the output of root script" exec > $LOG 2>&1 else mkfifo $LOG.pipe tee < $LOG.pipe $LOG & exec > $LOG.pipe 2>&1 rm $LOG.pipe fi Using a demonstration as an example of how exec works #!/bin/ksh LOGFILE="$0.log" PIPE="$0.pipe" [[ -e "$PIPE" ]] || mkfifo "$PIPE" # or mknod -p ?? # make a new channel(3) and copy channel 1's destination as its own (does NOT POINT TO channel 1's destination) # this allows the normal output to continue to STDOUT but also get printed to whatever file is attached to channel 3 exec 3>&1 # anything coming in on the pipe, send it to $LOGFILE and to channel 3 tee "$LOGFILE" <"$PIPE" >&3 & # redirect STDOUT to the pipe exec > "$PIPE" echo "going to default output" echo "forcing to channel 1" >&1 echo "forcing to channel 2" >&2 echo "forcing to channel 3" >&3 rm -f "$PIPE" ==== Another example of how to set up simultaneous output to both terminal and a logfile using file descriptors ==== exec 3<> /tmp/foo #open fd 3 for r/w echo "test" >&3 exec 3>&- #close fd 3. exec 3<> myfile.txt while read line <&3 do { echo "$line" (( Lines++ )); # Incremented values of this variable #+ accessible outside loop. # No subshell, no problem. } done exec 3>&- echo "Number of lines read = $Lines" # 8 Mmm. See our output and also tee it to a log file! #!/bin/bash echo hello if test -t 1; then # Stdout is a terminal. exec >log else # Stdout is not a terminal. npipe=/tmp/$$.tmp trap "rm -f $npipe" EXIT mknod $npipe p tee <$npipe log & exec 1>&- exec 1>$npipe fi echo goodbye === and another example === Found at [[https://www.unix.com/shell-programming-and-scripting/70726-how-redirect-script-output-inside-script.html]] #!/bin/bash #Objective : To redirect the stdout & stderr to two different files from within the script and without sacrificing the scripts output mkfifo /tmp/out.pipe /tmp/err.pipe exec 3>&1 4>&1 tee /tmp/std.out < /tmp/out.pipe >&3 & pid_out=$! exec 1>/tmp/out.pipe tee /tmp/std.err < /tmp/err.pipe >&4 & pid_err=$! exec 2>/tmp/err.pipe log() { echo "`date` : $@ " } log "Starting the process" log "Listing directories now..." ls -l ls -l Log_to_Stderr exec 1>&3 3>&- 2>&4 4>&- wait $pid_out wait $pid_err rm /tmp/out.pipe /tmp/err.pipe log "End of Processing" === More elaborate example === found [[http://stackoverflow.com/questions/2288939/create-a-pipe-that-writes-to-multiple-files-tee|here on stackoverflow]] #!/bin/sh # Author: Harvey Chapman # Description: POSIX shell functions that can be used with tee to simultaneously put # stderr and stdout to both a file and stdout # # Based on: # Re: How to redirect stderr and stdout to a file plus display at the same time # http://www.travishartwell.net/blog/2006/08/19_2220 # # Original example function from Travis Hartwell's blog. # Note: I've made minor changes to it. example() { OUTPUT_LOG=output.log OUTPUT_PIPE=output.pipe # This should really be -p to test that it's a pipe. if [[ ! -e $OUTPUT_PIPE ]]; then mkfifo $OUTPUT_PIPE fi # This should really be -f to test that it's a regular file. if [[ -e $OUTPUT_LOG ]]; then rm $OUTPUT_LOG fi exec 3>&1 4>&2 tee $OUTPUT_LOG < $OUTPUT_PIPE >&3 & tpid=$! exec > $OUTPUT_PIPE 2>&1 echo "This is on standard out" echo "This is on standard err" >&2 exec 1>&3 3>&- 2>&4 4>&- wait $tpid rm $OUTPUT_PIPE } # A slightly reduced version of example() example2() { OUTPUT_LOG=output.log OUTPUT_PIPE=output.pipe rm -f $OUTPUT_PIPE mkfifo $OUTPUT_PIPE rm -f $OUTPUT_LOG tee $OUTPUT_LOG < $OUTPUT_PIPE & tpid=$! exec 3>&1 4>&2 >$OUTPUT_PIPE 2>&1 echo "This is on standard out" echo "This is on standard err" >&2 exec 1>&3 3>&- 2>&4 4>&- wait $tpid rm -f $OUTPUT_PIPE } # # Logging methods based on above. See the example below for how to use them. # # Usage: start_logging [[delete_existing_logfile]] start_logging() { # Check to see if OUTPUT_LOG and OUTPUT_PIPE need to be defined. if [[ -z "$OUTPUT_LOG" ]]; then OUTPUT_LOG=output.log fi if [[ -z "$OUTPUT_PIPE" ]]; then OUTPUT_PIPE=output.pipe fi # Make sure that we're not already logging. if [[ -n "$OUTPUT_PID" ]]; then echo "Logging already started!" return 1 fi # Always remove the log and pipe first. rm -f $OUTPUT_PIPE # Delete the logfile first if told to. if [[ "$1" = delete_existing_logfile ]]; then rm -f $OUTPUT_LOG fi mkfifo $OUTPUT_PIPE tee -a $OUTPUT_LOG < $OUTPUT_PIPE & OUTPUT_PID=$! exec 3>&1 4>&2 >$OUTPUT_PIPE 2>&1 } stop_logging() { # Make sure that we're currently logging. if [[ -z "$OUTPUT_PID" ]]; then echo "Logging not yet started!" return 1 fi exec 1>&3 3>&- 2>&4 4>&- wait $OUTPUT_PID rm -f $OUTPUT_PIPE unset OUTPUT_PID } example3() { start_logging #start_logging delete_existing_logfile echo "This is on standard out" echo "This is on standard err" >&2 stop_logging } ==== Execute an SQL file from a shell and email the results to a user ==== This is bare-bones with no error-checking but works as tested.\\ The idea is to gather up the results throughout the day (however many times it is run from crontab) and the first time it is run on the following day, it will zip up the file from the previous day and mail to to a user. #!/usr/bin/ksh export PATH="/usr/local/bin:${PATH}" export ORACLE_SID="$1" ORAENV_ASK=NO . oraenv SPOOLDIR="/tmp" SPOOLSTUB="${ORACLE_SID}_pga_stats" SPOOLFILE="${SPOOLSTUB}_$(date '+%y%m%d').csv" sqlplus '/ as sysdba'< ==== RedHat root filesystem has gone read-only ==== * [[http://www.unixarena.com/2013/09/redhat-linux-how-to-fix-read-only-root.html|How to fix read only root file system]] ==== Kill all processes for a user ==== for prc in `ps -ef | grep -E "^ +[o]racle" | awk '{print $2}'`; do kill $prc done ==== isRGHere ==== Checks if resource group is on this leg of an HACMP cluster. Returns 0 if true else 1. #!/usr/bin/ksh SCRIPT=`basename $0` function rctest { exitcode=$1 msg=$2 if [[ $exitcode -ne 0 ]]; then echo "********************************************************************************" echo "; Script $SCRIPT finished with errors." echo "$msg." echo "Returncode : $exitcode." echo "; * ***************************************************************************" fi exit $exitcode } RGINFO=/usr/es/sbin/cluster/utilities/clRGinfo [[ ! -f $RGINFO ]] && rctest 1 "clRGinfo not found" if [[ $# -eq 1 ]] then RG=`echo $1 | cut -c 1-14` else rctest 10 "Usage: `basename $0` " fi $RGINFO |grep -qwp $RG || rctest 9 "$RG is not defined" THISNODE=`/usr/es/sbin/cluster/utilities/get_local_nodename | cut -c 1-14` $RGINFO |grep -wp $RG |grep -w $THISNODE |grep -wq ONLINE or this one-liner will work by returning ONLINE or OFFLINE or nothing in case of non-clustered machine /usr/es/sbin/cluster/utilities/clRGinfo|grep $(hostname|awk -F. '{print $1}')|awk '{print $(NF-1)}' ==== Find cluster name for current node of an hacmp cluster ==== /usr/es/sbin/cluster/utilities/cltopinfo|grep 'Cluster Name'|awk '{print $NF}' ==== AIX: add a user to a group ==== chgrpmem -m + oracle ibmtls ==== Find swear words in scripts by comparing it to a list found on the internet ==== wget "http://www.bannedwordlist.com/lists/swearWords.csv" -O /tmp/s ; for word in $(cat /tmp/s | sed -e 's/ /_/g' -e 's/,/ /g') ; do grep -wR $word *; done | less ==== grep -p (for paragraph) works on AIX but not on Linux or Solaris ==== Use awk instead awk 'BEGIN {FS="\n" RS="\n\n"} /search pattern/ { do something }' /usr/xpg4/bin/awk 'BEGIN {RS="\n\n";FS="\n"} /AGRHDWQ1/ {print $2}' dsm.sys | awk '{print $NF}' this prints the last word of the second //line// in the paragraph in dsm.sys that contains the search term AGRHDWQ1.\\ Or slightly simpler... awk -v RS='' '/NHAPPLP1/' /etc/tsm/dsm.sys # (use /usr/xpg4/bin/awk on Solaris) or, case insensitively: awk -v RS='' 'tolower($0) ~/so_u_clubqa_orx_d_cab/' /etc/tsm/dsm.sys Using -v means you don't have to use a BEGIN section. ==== debug/redirect log of a shell ==== #!/usr/bin/ksh exec 2>/tmp/mytest ==== Different ways of Iteration in korn shell with a while loop ==== IFS="|" exec 0<$statfile while read host db started stopped do rrdfile="export_duration_${host}_${db}.rrd" $RRDTOOL update ${RRDDIR}/${rrdfile} $started:$started $stopped done or while read host db started stopped do rrdfile="export_duration_${host}_${db}.rrd" $RRDTOOL update ${RRDDIR}/${rrdfile} $started:$started $stopped done or cat $statfile | sort -nk4 | while IFS=\| read host db type started stopped do [[ "$stopped" == "" ]] && continue rrdfile="export_duration_${host}_${db}.rrd" $RRDTOOL update ${RRDDIR}/${rrdfile} ${started}:${started}:${stopped} [[ $? -ne 0 ]] && echo "nok: $?" done ==== Is it a directory or a filesystem mountpoint? ==== for i in /cln/exp/ora_data3/*; do FS="" MNT="$(df -g $i|grep -v Filesystem|awk '{print $NF}')" if [[ "$i" == "${MNT}" ]]; then FS=" (FS)" fi echo "${i}${FS}" done ==== Filesystem 100% full, what's taking up all the space? ==== find /oracle/endur -xdev -ls|sort -nr +6|head or /dev/esb01fs010001 1.00 0.00 100% 1838 69% /oracle cd /oracle du -gsx * | sort -n 0.00 checkpoints 0.00 flash_recovery_area 0.00 lost+found 0.00 oraInst.loc 0.00 oraInventory 0.09 admin 0.99 diag cd diag and repeat until culprits are found ==== df -g shows filesystem full but du -gsx shows it as not full! ==== Probably caused by some large files being deleted while a process was still writing to them (and is continuing to write to them). Solution is either kill the process that was writing to the file (descriptor) or find the file descriptors associated with the removed files and close them. This lists all the files under that filesystem that have been deleted (unlinked) lsof +aL1 /cln/tst sort -nk 7 | tail -10 Then look through the /proc filesystem for the file descriptors linked to the removed files ls -ltr /proc//fd | grep Then we need to use a debugger to close the descriptors dbg -p call close () quit ==== Capitalise the first letter of a word in Korn Shell ==== Not as easy as it sounds if it needs to work in non-ksh93 shells function capit { typeset -u first first=${1%"${1#?}"} printf "%s\n" "${first}${1#?}" return 0 } or maybe more neatly done in Perl...\\ This will capitalise each word in a sentence passed into it function capit { echo "$1" | perl -pe "s/([\w']+)/\u\L\1/g" return 0 } ==== Remove / rename a directory / file containing weird control characters ==== Use ls with -i to see inode listing ls -bali Use find with -inum to get the filename and -exec to remove it find . -inum -exec rm -f {} \; ==== Run a local script on a remote host ==== Doing it this way means we do not need to scp the shell over to the host before executing it! ssh user@host 'sh' < your_script.sh First of all, this command is a redirection: your shell will open the file your_script.sh and feed it as input to the ssh command. ssh, in turn, will tunnel its stdin to the remote command, namely, sh instance. And sh without arguments reads its script from stdin. So we got sh instance, which is launched on remote host, but reads commands from your local file. ==== Get a list of running instances - but only those started by the current user ==== This is the way I wanted to do it but there's an issue. Where does that spare } character come from? Anyone? ps -ef | grep [p]mon | awk -v dbowner=$(id -un) '{ if ($1==dbowner) { gsub(/ora_pmon_/,"",$NF); print $NF; } }' ps listing $ ps -ef | grep [p]mon oracle 13304024 1 0 Jun 07 # 2:39 ora_pmon_reportk oracle 26018178 1 0 Jun 07 # 3:01 ora_pmon_dwh_perf oracle 29229432 1 0 Jun 07 # 2:30 ora_pmon_adso oraebso 18022994 1 0 Jun 07 # 2:38 ora_pmon_EBSO oracle 30278192 1 0 Jun 07 # 2:48 ora_pmon_owb112k Results of above command reportk dwh_perf adso } owb112k Workaround 1. Send the ps listing to a file and work on it without pipes. Works but it's a bit long-winded especially as we have to clean up afterwards. ps -ef | grep [p]mon>/tmp/results.$$ 2>/dev/null; awk -v dbowner=$(id -un) '{ if ($1==dbowner) { gsub(/ora_pmon_/,"",$NF); print $NF; } }' /tmp/results.$$; rm -f /tmp/results.$$ Workaround 2. Don't like the grep -v but until I find out where that } is coming from.. ps -ef | grep [p]mon | awk -v dbowner=$(id -un) '{ if ($1==dbowner) { gsub(/ora_pmon_/,"",$NF); print $NF; } }' | egrep -v 'grep|}|ASM' Found it. ps is particular. Cannot use ps -ef as the input to commands like awk. Do it like this... ps -ef | grep [p]mon|while read line; do awk -v dbowner=$(id -un) '{ if ($1==dbowner) { gsub(/ora_pmon_/,"",$NF); print $NF; } }';done Also see [[http://mywiki.wooledge.org/BashPitfalls#for_f_in_.24.28ls_.2A.mp3.29|bash pitfalls]] ==== Return a list of Oracle databases running on a list of remote servers ==== #!/bin/ksh serverlist=`cat /home/tools/ini/system/oracle_servers | sort -n | tr "\n" " "` if [ -z "${serverlist}" ] then echo "no servers found" && exit 1 fi for server in ${serverlist} do ssh ${server} "ps -ef | grep [o]ra_pmon_" >/tmp/${server}.dblist done for server in ${serverlist} do cat /tmp/${server}.dblist | awk -F_ -v SRV=${server} 'BEGIN {print SRV ":"} {print $NF}' | tr "\n" " " echo done ==== Clever trick to check whether SQL output needs sending to someone ==== Using an exit status from SQL*Plus back to the shell so we can decide what to do. Nice one! #!/bin/bash tempfile=/tmp/audit_locked_accounts_$ORACLE_SID.txt # Start sqlplus and check for locked accounts sqlplus -S "/ as sysdba" << EOF > $tempfile set pages 0 select 'The following accounts were found to be unlocked and should not be' from dual; def exit_status = 0 col xs new_value exit_status select username , account_status , 1 as xs from dba_users where account_status != 'LOCKED' and username in ('HR', 'SCOTT', 'OUTLN', 'MDSYS', 'CTXSYS') / exit &exit_status EOF # If the exit status of sqlplus was not 0 then we will send an email if [[ $? != 0 ]]; then mail -s "Accounts Unlocked in $ORACLE_SID" oracle < $tempfile fi ==== Check RMAN logfiles for errors from cron every day ==== 00 09 * * * /home/tools/scripts/oracle/dosh 'find /home/tools/logs/rman -name "*online.log" -mtime -1 -exec sed -ne "/^RMAN-/,/^$/p" {} \\; -ls' | mailx -s 'RMAN errors last night' stuart@domain.com /home/ibmtools/scripts/oracle/dosh 'find /oracle/export -name "expdp*log" -mtime -1 -exec grep ORA- {} \; -ls' | mailx -s 'Datapump errors for Baxter last night' stuart@domain.com ==== Standard getopts ==== Only allows single character parameter keys. while getopts "vf:" flag do case "$flag" in v) VERBOSE=1;; f) FILE2RUN=$OPTARG;; esac done shift $((OPTIND-1)) ==== An alternative to getopts ==== These methods have the advantage of allowing arguments of any length not just 1 character # Parse command-line arguments INST_ARG= DO= while [ $# -gt 0 ] do case $1 in -force) FORCE=yes shift ;; -vm) echo echo "*** The \"-vm\" option has been deprecated." echo "***" echo "*** Now there is only one set of services running for all virtual machines." echo "*** If you meant to address a specific installation *instance*, please" echo "*** use the \"-instance\" option. If you meant to do smth to services running on" echo "*** a specific virtual machine, then don't pass any option at all." echo exit 1 ;; -instance|-focus) shift if [ $# -eq 0 ] then echo echo "*** Instance name expected after \"-instance\" option." echo exit 1 fi INST_ARG=$1 shift ;; -all) INST_ARG=all shift ;; add|-add|+) DO="add" shift PKGIDS="$1" [ x$PKGIDS = x ] || shift ;; start|-start) DO="start" shift ;; stop|-stop) DO="stop" shift ;; restart|-restart) DO="restart" shift gui@*|-gui@*) GUI_CS="`echo $1 | awk -F\@ '{print $2}'`" DO="gui" shift ;; registry|-registry|reg|-reg) DO="registry" shift ;; -fuse) shift if [ $# -eq 0 ] then echo echo "*** Instance name expected after \"-fuse\" option." echo exit 1 fi FUSE_INST_ARG=$1 shift ;; shell|-shell) DO="shell" shift ;; direct|-direct|systemctl) NO_REDIRECT_TO_SYSTEMCTL="YES" shift ;; *) echo echo "*** Unknown argument \"$1\" encountered." echo exit 1 ;; esac done another one... countArgs=$# while [[ $# -gt 0 ]] do idx1=$($ECHO $1|$GREP "="|wc -l) if [[ $idx1 -gt 0 ]]; then key=$($ECHO $1|$CUT -d '=' -f 1) value=$($ECHO $1|$CUT -d '=' -f 2) else key=$($ECHO $1|$CUT -d '=' -f 1) idx2=$($ECHO $1|$GREP "^-"|wc -l) if [[ $idx2 -eq 0 ]]; then $ECHO -e "\n ERROR: $1 is an unsupported argument passed to agentDeploy.sh.\n" usage exit 1 fi fi case "$key" in -help) if [[ $upgradeFlag ]]; then upgradeUsage else freshUsage fi exit 0;; AGENT_BASE_DIR) agentBaseDir=$($ECHO $value|$SED 's/\/$//') checkBaseFlag=TRUE;; OMS_HOST) omsHost=$value checkOMSHost=TRUE;; EM_UPLOAD_PORT) omsPort=$value checkOMSPort=TRUE;; AGENT_INSTANCE_HOME) instHome=$($ECHO $value | $SED 's/\/$//');; AGENT_REGISTRATION_PASSWORD) pswd=$value;; s_encrSecurePwd) pswd=$value;; RESPONSE_FILE) rspFlag=TRUE rspLoc=$value;; OLD_AGENT_VERSION) oldAgentVersion=$value;; OLDHOME) oldHome=$value;; -debug) debugSwitch=TRUE;; -forceConfigure) forceFlag=TRUE;; -configOnly) configFlag=TRUE validationFlag=TRUE;; -agentImageLocation) archiveLoc=$value checkArchiveFlag=TRUE;; -invPtrLoc) shift ptrLoc=$1;; -Upgrade) UpgradeFlag=TRUE validFlag=TRUE;; INVENTORY_LOCATION) validFlag=TRUE;; b_forceInstCheck) validFlag=TRUE forcefullFlag=TRUE;; -prereqOnly) validationFlag=TRUE prereqFlag=TRUE;; -executeRootsh) validFlag=TRUE;; *) idx=$($ECHO $1|$GREP "^-"|wc -l) if [[ $idx -eq 0 ]]; then validFlag=TRUE else $ECHO -e "\n ERROR: Invalid argument $key passed." usage exit 1 fi esac shift done ==== getopts - another way - found in /etc/init.d/functions (daemon function) in Fedora ==== while [ "$1" != "${1##[-+]}" ]; do case $1 in '') echo $"$0: Usage: daemon [+/-nicelevel] {program}" "[arg1]..." return 1 ;; --check) base=$2 gotbase="yes" shift 2 ;; --check=?*) base=${1#--check=} gotbase="yes" shift ;; --user) user=$2 shift 2 ;; --user=?*) user=${1#--user=} shift ;; --pidfile) pid_file=$2 shift 2 ;; --pidfile=?*) pid_file=${1#--pidfile=} shift ;; --force) force="force" shift ;; [-+][0-9]*) nice="nice -n $1" shift ;; *) echo $"$0: Usage: daemon [+/-nicelevel] {program}" "[arg1]..." return 1 ;; esac done ==== getopts - yet another way - found in adstrtal.sh (middleware start script for EBS) ==== # # Parse Arguments # for nxtarg in $* do arg=`echo $nxtarg | sed 's/^//'` case $arg in -secureapps) if test "$secureapps" = "" ; then secureapps=$arg else echo "$0: Duplicate Argument passed : $arg" usage fi ;; -nodbchk) if test "$nodbchk" = "" ; then nodbchk=$arg else echo "$0: Duplicate Argument passed : $arg" usage fi ;; -nopromptmsg) if test "$nopromptmsg" = "" ; then nopromptmsg=$arg else echo "$0: Duplicate Argument passed : $arg" usage fi ;; *) if test "$unpw" = "" ; then unpw=$arg else echo "$0: Duplicate Argument passed : $arg" usage fi esac done ==== getopts - and another in adautocfg.sh ==== for myarg in $* do arg=`echo $myarg | sed 's/^-//'` case $arg in appspass=*) appspass=`echo $arg | sed 's/appspass=//g'` shift ;; nocustom) myparams="$myparams $arg" shift ;; noversionchecks) myparams="$myparams $arg" shift ;; promptmsg=*) promptmsg=`echo $arg | sed 's/promptmsg=//g'` shift ;; *) echo "$0: unrecognized action specified" exit 1 esac done ==== getopts - ok, this is the last one (from ${OMS_HOME}/OMSPatcher/wlskeys/createkeys.sh ==== ### parse ### oh=${ORACLE_HOME} location= getoh=0 getloc=0 gethelp=0 for arg in $@ do if [ $getoh = 1 ]; then oh=$arg getoh=2 fi if [ $getloc = 1 ]; then location=$arg getloc=2 fi if [ "$arg" = "-oh" ]; then getoh=1 fi if [ "$arg" = "-location" ]; then getloc=1 fi if [ "$arg" = "-help" ]; then gethelp=1 fi done ### help ### if [ $gethelp = 1 ] || [ -z "${oh}" -o -z "${location}" ]; then echo "Usage:" echo " ${binname} [-help]" echo " ${binname} -oh -location " echo "" echo " -oh OMS Platform home path" echo " -location Location to store the configuration and key files" exit 1 fi ==== Run a script on multiple servers ==== #!/usr/bin/env ksh # ============================================================================== # Name : dosh (Distributed Oracle SHell) # Description : Runs a command on all Oracle servers # # Parameters : h - displays help # b - execute in the background (in parallel) # v - verbose (default, like all unix commands is silent) # c - command to be executed # f - file containing commands to be executed # # Example : ./dosh -v -c 'ls -al' # ./dosh -v -f /tmp/complex_shell # # Notes : Escape " and $ with \ # # Modification History # ==================== # When Who What # ========= ================= ================================================== # 08-FEB-13 Stuart Barkley Created # 08-JUL-15 Stuart Barkley Background execution mode # ============================================================================== # -------------------------------------------------------------------------- # this list of servers is generated by the DBAHAWK tool so will be up to date # -------------------------------------------------------------------------- serverfile=/oracle/scripts/.shb/oracle_servers [[ -z "${serverfile}" ]] && echo "Server list ${serverfile} not found" && exit 1 serverlist=`cat ${serverfile} | sort -n | tr "\n" " "` # ------------------------- # get the arguments, if any # ------------------------- while getopts "hbvc:f:" OPT do case "$OPT" in h) echo "\nUsage: $0 [-b] [-v] [-h] -c ''\n"; exit; ;; b) BACKGROUND="Y"; ;; v) VERBOSE="Y"; ;; c) CMMND=$OPTARG; ;; f) CMDFILE=$OPTARG; ;; *) echo "\nUsage: $0 [-b] [-v] [-h] [-c '' | -f ]\n"; exit; ;; esac done shift $((OPTIND-1)) VERBOSE=${VERBOSE:-"N"} # -------------------------------- # check we have required arguments # -------------------------------- [[ -z $CMMND && -z $CMDFILE ]] && printf "%s\n" "Usage: $0 [-b] [-v] [-h] [-c '' | -f ]" && exit 1 # --------------------------------- # put the thing to be run in a file # --------------------------------- if [[ ! -z $CMMND ]]; then printf "%s\n" "$CMMND" > /tmp/dosh.$$ else cp $CMDFILE /tmp/dosh.$$ fi # ---------------------------- # loop over the remote servers # ---------------------------- for server in ${serverlist} do if [[ "$VERBOSE" == "Y" ]]; then printf "%s %s\n" "${server}" "$(date '+%d-%b-%Y %H:%M:%S')" printf "%s\n" "-------------------------------------------" fi scp -q /tmp/dosh.$$ ${server}:/tmp/ [[ $? -ne 0 ]] && echo "WARNING: Failed to connect to ${server}. Are ssh keys setup?, skipping this server" && continue if [[ "$BACKGROUND" == "Y" ]]; then ssh -n -o BatchMode=yes ${server} "ksh /tmp/dosh.$$; rm -f /tmp/dosh.$$"> /tmp/dosh_${server}.out 2>&1 & else ssh -n -o BatchMode=yes ${server} "ksh /tmp/dosh.$$; rm -f /tmp/dosh.$$" fi [[ "$VERBOSE" == "Y" ]] && printf "\n\n" done rm -f /tmp/dosh.$$ [[ "$BACKGROUND" == "Y" ]] && printf "%s\n" "Type 'ls -altr /tmp/dosh*out' to see output from the commands" ==== Get a stack trace a running process on Linux ==== pstack ==== Get a stack trace a running process on AIX ==== procstack ==== Trace a running process on Linux ==== strace -frT -o /tmp/strace-output.$$ -p ==== Trace a running process on AIX ==== truss -rall -wall -xall -sall -p or truss -aefdDo /tmp/truss_output_$$ -p === Get AIX info ==== Borrowed from /opt/commvault/Base/get_sys_info_AIX.sh #!/bin/ksh CVREG="/etc/CommVaultRegistry" INSTANCE="Instance001" BASEDIR=`grep -w dBASEHOME $CVREG/Galaxy/$INSTANCE/Base/.properties | awk '{print $2}'` . $BASEDIR/common_info_funcs.sh print_header echo "List of Mountpoints" mount 2>&1 print_line echo "System Information (lsconf o/p)" lsconf 2>&1 print_line echo "Disk Information" lsdev -Cc disk 2>&1 print_line for disk in `lsdev -Cc disk -r name` do echo "Attributes for $disk" lsattr -El $disk 2>&1 print_line echo "lscfg -vl $disk" lscfg -vl $disk 2>&1 print_line done echo "LV Information" lsdev -Cc 'logical_volume' -s lvsubclass 2>&1 print_line for lv in `lsdev -Cc 'logical_volume' -s lvsubclass -r name` do echo "Detailed info for $lv" lslv $lv 2>&1 print_line echo "Attributes for $lv" lsattr -El $lv 2>&1 print_line echo "PV mapping for $lv" lslv -m $lv print_line done exit 0 ==== Find hardware information on AIX ==== lparstat -i CPU, memory and VG info etc. lsconf or a quicker way to find the Machine serial number... odmget CuAt | grep -p systemid | awk '$1 == "value" { print substr($3,8,7) }' ==== How can a shell script find out what directory it is in? ==== basename $0 does not always give the desired answer (might give "./")! DIR="$(cd "$(dirname "$0")" && pwd)" ==== Array processing ==== array=(1 2 3) unset array[2] echo ${array[2]} # null indices=(${!array[@]}) # create an array of the indices of "array" size=${#indices[@]} # the size of "array" is the number of indices into it size=${#array[@]} # same echo ${array[@]: -1} # you can use slices to get array elements, -1 is the last one, etc. for element in ${array[@]}; do # iterate over the array without an index for index in ${indices[@]} # iterate over the array WITH an index do echo "Index: ${index}, Element: ${array[index]}" done for index in ${!array[@]} # iterate over the array WITH an index, directly array+=("new element") # append a new element without referring to an index ((counter++)) # shorter than ((counter=counter+1)) or ((counter+=1)) if [[ $var == 3 ]] # you can use the more "natural" comparison operators inside double square brackets while [[ $var < 11 ]] # another example echo ${array[${index}-1] # math inside an array subscript ==== Send an email by talking directly to an smtp server ==== #!/bin/bash telnet smtp.domain.com 25 <>/tmp/smtp.log HELO me.domain.com MAIL FROM: RCPT TO: DATA From: Stuart To: Anne Subject: testing smtp email Hello, this should appear in the body . QUIT EOTXT ==== Send an email by talking directly to an smtp server via file descriptor (no telnet! this time), adding authentication ==== #!/bin/bash # # mail.sh # # 2008 - Mike Golvach - eggi@comcast.net # 2010 - Rayber # # Creative Commons Attribution-Noncommercial-Share Alike 3.0 United States License # if [ $# -ne 7 ] then echo "Usage: $0 FromAdress ToAdress Domain MailServer MailTextFile AuthEmail AuthPass" exit 1 fi from=$1 to=$2 domain=$3 mailserver=$4 mailtext=$5 authemail=`echo $6|openssl enc -base64|awk 'sub("..$", "")'` authpass=`echo $7|openssl enc -base64|awk 'sub("..$", "")'` if [ ! -f $mailtext ] then echo "Cannot find your mail text file. Exiting..." exit 1 fi exec 9<>/dev/tcp/$mailserver/25 echo "HELO $domain" >&9 read -r temp <&9 echo "$temp" echo "auth login" >&9 read -r temp <&9 echo "$authemail" >&9 read -r temp <&9 echo "$authpass" >&9 read -r temp <&9 echo "Mail From: $from" >&9 read -r temp <&9 echo "$temp" echo "Rcpt To: $to" >&9 read -r temp <&9 echo "$temp" echo "Data" >&9 read -r temp <&9 echo "$temp" cat $mailtext >&9 echo "." >&9 read -r temp <&9 echo "$temp" echo "quit" >&9 read -r temp <&9 echo "$temp" 9>&- 9<&- echo "All Done Sending Email. See above for errors" exit 0 ==== Send an html file to an email recipient from a Unox server using sendmail ==== ( echo "From: $(id -un)@$(hostname)"; echo "To: admin@company.com"; echo "Subject: Your HTML report"; echo "Content-Type: text/html"; echo "MIME-Version: 1.0"; cat /file/with/html/tags.html; ) | sendmail -t ==== rsync examples ==== * [[Distribute files to multiple servers using rsync and ssh]] * [[https://www.tecmint.com/rsync-local-remote-file-synchronization-commands/|rsync practical example - tecmint]] #!/bin/sh ssh '/bin/mkdir -p /etc /etc/rc.config.d /etc/security /etc/mail' rsync --rsync-path /usr/bin/rsync -Liprogtz --out-format=%f%L /etc/passwd /etc/passwd.post /etc/group /etc/hosts /etc/services /etc/resolv.conf /etc/exclude.rootvg :/etc rsync --rsync-path /usr/bin/rsync -Liprogtz --out-format=%f%L /etc/hosts.allow.xcat :/etc/hosts.allow rsync --rsync-path /usr/bin/rsync -Liprogtz --out-format=%f%L /etc/rc.config.d/sap :/etc/rc.config.d rsync --rsync-path /usr/bin/rsync -Liprogtz --out-format=%f%L /etc/security/group /etc/security/limits /etc/security/login.cfg /etc/security/passwd /etc/security/user :/etc/security rsync --rsync-path /usr/bin/rsync -Liprogtz --out-format=%f%L /etc/mail/sendmail.cf :/etc/mail rsync -av --progress /home/ibmtools/scripts/oracle/* benouerp07:/home/ibmtools/scripts/oracle/ rsync -avzh --progress --exclude='*dmp' --exclude='*csv' oracle@hn481:/oracle/scripts /Users/stuartb/oracle/ ==== Handy aliases ==== Strip out comments and blank lines from a file alias strip='grep -Ev '\''^(#|$)'\''' Does a ps and a grep alias psg='ps -ef | grep -v $$ | grep -i ' Getting a decent listing of filesystem space available. It is ordered such that the filesystems with no space left are at the end. OS=$(uname -s) [[ "$OS" == "SunOS" ]] && alias dfg='df -h|sed -e '1d'|sort -n -k5|awk '\'' BEGIN {printf("%-35s%-10s%-10s%-6s%-30s\n","Filesystem","Total","Free","%Used","Mounted on")} {printf("%-35s%-10s%-10s%-6s%-30s\n",$1,$2,$4,$5,$6)}'\''' [[ "$OS" == "AIX" ]] && alias dfg='df -g|sed -e '1d'|sort -n -k4|awk '\'' BEGIN {printf("%-35s%-10s%-10s%-6s%-30s\n","Filesystem","Total GB","Free","%Used","Mounted on")} {printf("%-35s%-10s%-10s%-6s%-30s\n",$1,$2,$3,$4,$7)}'\''' A handy way of listing subdirectories and their files alias filetree="find . -print | sed -e 's;[^/]*/;|____;g;s;____|; |;g'" Watch progress of a copy alias cpProgress="rsync --progress -ravz" Reboots Linksys router alias rebootlinksys="curl -u 'admin:password' 'http://192.168.1.2/setup.cgi?todo=reboot'" Nice one for bash. Colour codes the prompt depending on the outcome of the previous command bash_prompt_command() { RTN=$? prevCmd=$(prevCmd $RTN) } PROMPT_COMMAND=bash_prompt_command prevCmd() { if [ $1 == 0 ] ; then echo $GREEN else echo $RED fi } if [ $(tput colors) -gt 0 ] ; then RED=$(tput setaf 1) GREEN=$(tput setaf 2) RST=$(tput op) fi export PS1="\[\e[36m\]\u.\h.\W\[\e[0m\]\[\$prevCmd\]>\[$RST\]" Mmmm, to be looked into. Executes remote commands on a unix box using curl. #/bin/sh # # WAG320N-HACK # Ver. 1.0 # 12/09/2010 # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # # You should have received a copy of the GNU General Public License # along with this program. If not, see . # Set username and password in the form of "username:password" # example: "admin:admin" my_access="admin:admin" # Parameters test if [ -z "$1" ] then echo "wag320n-hack.sh: missing remote command" echo "usage: wag320n-hack.sh " echo "example: wag320n-hack.sh /bin/ls -la /usr/sbin" echo "Note: always use full path" echo "" echo "wag320n-hack.sh - Ver. 1.0 - 12/09/2010" echo "Licensed under GPL V. 3" echo "" exit 0 fi # Get the command my_command="ping_size="'$('"$@"' 1>&2)' curl -s -G -u "$my_access" --data-urlencode 'todo=ping_test' --data-urlencode 'this_file=Diagnostics.htm' --data-urlencode 'next_file=Ping.htm' --data-urlencode 'c4_ping_ipaddr=192.168.1.1' --data-urlencode 'ping_timeout=5000' --data-urlencode 'ping_interval=1000' --data-urlencode 'ping_number=1' --data-urlencode "$my_command" http://192.168.1.1/setup.cgi | sed -e '/HTTP\/1.0 200 OK/q' ==== How to configure SSH with public/private keys ==== * [[http://www.unixpeople.com/HOWTO/configuring.ssh.html|unixpeople.com]] ==== Use SSH config file ==== Host server10 Hostname 1.2.3.4 IdentityFile ~/.ssh/id_dsa user foobar Port 30000 ForwardX11Trusted yes TCPKeepAlive yes then just connect using ssh server10 ==== A decent sed tutorial ==== From [[http://www.grymoire.com/Unix/Sed.html|grymoire.com]] ==== A decent korn/bash shell tutorial ==== From [[http://www.dartmouth.edu/~rc/classes/ksh/print_pages.shtml|dartmouth.edu]] Reproduced here just in case it disappears! [[Advanced shell scripting]] ==== trap ==== Example Handling Traps With ksh - Discussion of the kill command EXAMPLE TEMPLATE: PRODUCT: HP-UX 11iV1 Version B.11.11 HP Tru64 V5.1B PK4 Sun/Solaris SunOS V5.8 Linux 2.6 kernel COMPONENT: ksh SOURCE: Philippe Vouters Fontainebleau/France LOW COST HIGH-TECH PRODUCTS: http://techno-star.fr OVERVIEW: The ksh script below shows how to eventually handle traps in the situation where someone might try to kill a script by killing individual commands run by that script or the entire process group a script is running in. The kill command (usually a shell builtin) may be used to send a signal to a process group (with the - syntax) or an individual process. The example ksh script below runs /bin/sleep as the foreground process, the example ksh scripts immediately returns when the /bin/sleep process has terminated. Most signals sent to the shell are ignored until after the foreground process terminates. This is in order to avoid creating zombie processes. Therefore a kill on the example ksh script waits for the termination of the /bin/sleep process. The status value $? in the trap refers to the exit status of the command to run and therefore is the exit status of the /bin/sleep process. The called function in the trap handler shows how to correctly examine the effect of the kill command on the shell or it's children. To examine the value of $? in a trap handler means that you must understand what it can be set and how different signals delivered to either the shell or the foreground process (or the process group) might affect the value of $?. The example shell script prints $? using echo but it does not perform tests on the value of $?. For a complete solution when attempting to trap signals in a shell you would also need code that examined the value of $? after the foreground process had completed. * CAUTION *** This sample script has been tested using HP-UX B.11.11, HP Tru64 V5.1B PK4, SunOS V5.8 and Fedora Core 4 (homed version of Red Hat Linux). However, we cannot guarantee its effectiveness because of the possibility of error in transmitting or implementing it. It is meant to be used as a template for writing your own scripts, and may require modification for use on your system. SCRIPT NOTES: To notice that the ksh script and /bin/sleep share the same process group identifier (PGID), issue the following commands: [philippe@victor ~]$ who philippe :0 Jan 10 10:16 philippe pts/1 Jan 10 21:30 (:0.0) philippe pts/2 Jan 10 21:30 (:0.0) [philippe@victor ~]$ tty /dev/pts/1 [philippe@victor ~]$ ps -j -t pts/2 PID PGID SID TTY TIME CMD 11072 11072 11072 pts/2 00:00:00 bash 11113 11113 11072 pts/2 00:00:00 ksh 11116 11113 11072 pts/2 00:00:00 sleep In this case sending kill -INT -11113 will send SIGINT to the process group 11113. Both of the ksh and sleep processes are contained within this process group. Important Note: On HP-UX, you have to $ export UNIX95=1 in order to be able to use the -j option of the ps command. SCRIPT: COPYRIGHT (C) 2005 BY HEWLETT-PACKARD COMPANY ALL RIGHTS RESERVED. THIS SOFTWARE IS FURNISHED UNDER A LICENSE AND MAY BE USED AND COPIED ONLY IN ACCORDANCE WITH THE TERMS OF SUCH LICENSE AND WITH THE INCLUSION OF THE ABOVE COPYRIGHT NOTICE. THIS SOFTWARE OR ANY OTHER COPIES THEREOF MAY NOT BE PROVIDED OR OTHERWISE MADE AVAILABLE TO ANY OTHER PERSON. NO TITLE TO AND OWNERSHIP OF THE SOFTWARE IS HEREBY TRANSFERRED. THE INFORMATION IN THIS SOFTWARE IS SUBJECT TO CHANGE WITHOUT NOTICE AND SHOULD NOT BE CONSTRUED AS A COMMITMENT BY HEWLETT-PACKARD COMPANY. HP ASSUMES NO RESPONSIBILITY FOR THE USE OR RELIABILITY OF ITS SOFTWARE ON EQUIPMENT THAT IS NOT SUPPLIED BY HP. NO RESPONSIBILITY IS ASSUMED FOR THE USE OR RELIABILITY OF SOFTWARE ON EQUIPMENT THAT IS NOT SUPPLIED BY HEWLETT-PACKARD COMPANY. SUPPORT FOR THIS SOFTWARE IS NOT COVERED UNDER ANY HP SOFTWARE PRODUCT SUPPORT CONTRACT, BUT MAY BE PROVIDED UNDER THE TERMS OF THE CONSULTING AGREEMENT UNDER WHICH THIS SOFTWARE WAS DEVELOPED. #!/bin/ksh function handle_signal { print -n "pid $$ recieved $2 " if [[ $1 = 0 ]];then print but foreground command ended successfully else if [[ $1 = $3 ]];then print and so did the last foreground command else print -n "and the exit status of the last foreground " print command was $1 fi fi # Kill our process group and then ourselves with SIGTERM, giving a # pid of 0 sends the signal to our process group. Killing the process # group should kill us as well, this assumes that SIGTERM is not # handled by any process in the process group. # # This code could be replaced with an exit with an exit value that # would indicate what the problem was to the caller. That is replace # these two lines with: # # exit $3 # # or a specific exit code could be used. # kill -TERM 0 kill -TERM $$ } OS=$(uname -a | awk '{print $1}') if [[ "$OS" = "Linux" ]]; then offset=256 elif [[ ("$OS" = "HP-UX") || ("$OS" = "SunOS") || ("$OS" = "OSF1") ]]; then offset=128 fi trap 'RC=$?; handle_signal $RC SIGINT $offset+2' INT trap 'RC=$?; handle_signal $RC SIGQUIT $offset+3' QUIT /bin/sleep 20 echo $? ==== DNS not working ==== Ping to an IP address works ping 74.125.136.103 but this doesn't ping www.google.com Check resolv.conf cat /etc/resolv.conf nameserver 95.130.132.17 nameserver 95.130.132.18 I had changed internet provider and forgot to update this. Just to set it to the router address and let that do the resolution nameserver 192.168.1.1 ==== Create new image with kvm ==== Ref: http://www.cyberciti.biz/faq/kvm-virtualization-in-redhat-centos-scientific-linux-6/\\ Build an empty space for a CentOS virtual machine qemu-img create -f qcow2 centos.img 12G Tried creating image with sudo virt-install -n CentOS --description "Trying out CentOS" --ram=1024 --vcpus=1 --cpu host --hvm --cdrom /home/bey9at77/Downloads/c6-x86_64-20130910-1.qcow2 --graphics vnc --disk path=/var/lib/libvirt/images/centos.img,bus=virtio,size=10 gives error Starting install... Allocating 'centos.img' | 10 GB 00:00 ERROR internal error Process exited while reading console log output: char device redirected to /dev/pts/1 qemu-kvm: -drive file=/home/bey9at77/Downloads/c6-x86_64-20130910-1.qcow2.bz2,if=none,media=cdrom,id=drive-ide0-1-0,readonly=on,format=raw: could not open disk image /home/bey9at77/Downloads/c6-x86_64-20130910-1.qcow2: Permission denied ==== rpm / yum commands ==== === Install an RPM Package === rpm -ivh pidgin-2.7.9-5.el6.2.i686.rpm -i - install -v - verbose -h - print progress hashes === Check dependencies of RPM Package before Installing === rpm -qpR BitTorrent-5.2.2-1-Python2.4.noarch.rpm -q - query -p - list package capabilities -R - list dependent package capabilities === Install RPM Package with all dependencies === yum install BitTorrent-5.2.2-1-Python2.4.noarch.rpm === Install RPM Package with all dependencies (when RPM has been downloaded to local machine) === yum localinstall BitTorrent-5.2.2-1-Python2.4.noarch.rpm === Force Install a RPM Package without dependencies === Package will not work if dependencies are required rpm -ivh --nodeps BitTorrent-5.2.2-1-Python2.4.noarch.rpm === Check an Installed RPM Package === rpm -q BitTorrent === List all files of an installed RPM package === rpm -ql BitTorrent === List All Installed RPM Packages === rpm -qa === Query information about an installed RPM package === rpm -qi vsftpd === Query information about a not yet installed RPM package === rpm -qip sqlbuddy-1.3.3-1.noarch.rpm === (Forcibly) Remove an RPM Package === Use package name (as seen in -qi above), not full name rpm -ev (--nodeps) vsftpd === Query a file that was installed as part of an RPM Package (which package contained this file) === rpm -qf /usr/bin/htpasswd === Verify an RPM package === Compares information of installed files of the package against the rpm database rpm -Vp sqlbuddy-1.3.3-1.noarch.rpm === Rebuild corrupted RPM database === cd /var/lib rm __db* rpm --rebuilddb rpmdb_verify Packages ==== Install rpmforge repository ==== * Download rpmforge-release-0.5.3-1.el6.rf.x86_64.rpm * Import the key sudo rpm --import http://apt.sw.be/RPM-GPG-KEY.dag.txt * Install the repository sudo rpm -i rpmforge-release-0.5.3-1.el6.rf.x86_64.rpm * Check the installation rpm -K rpmforge-release-0.5.3-1.el6.rf.x86_64.rpm * Test it sudo yum install terminator ==== Install rpmfusion repository ==== su -c 'yum localinstall --nogpgcheck http://download1.rpmfusion.org/free/el/updates/6/i386/rpmfusion-free-release-6-1.noarch.rpm http://download1.rpmfusion.org/nonfree/el/updates/6/i386/rpmfusion-nonfree-release-6-1.noarch.rpm' ==== config file for yum ==== Checkout this file for global yum config /etc/sysconfig/yum-cron-background ==== Setup Oracle Enterprise Linux (RedHat) with yum server ==== You need to download the yum .repo file from the server, as per the steps below. After this, you need to enable a flag in the .repo file as per your operating system version. Having done these two steps, when you run yum install command on your linux box, the Oracle's yum server will be scanned, the dependent & the relevant rpm's will be download and installed for you.\\ cd /etc/yum.repos.d To download files here wget http://public-yum.oracle.com/public-yum-el5.repo A file named public-yum-el5.repo will be created in your directory\\ Edit this file and enter enabled=1 against the operating systems which is relevant to you vi public-yum-el5.repo Next run the yum command yum install package-name ==== To change to static IP address (Raspberry Pi) ==== As root: cd /etc/networks vi interfaces replace the line “iface eth0 inet dhcp” with iface eth0 inet static address 192.168.1.100 netmask 255.255.255.0 gateway 192.168.1.1 You should also take a look at the file /etc/resolv.conf and check it has a nameserver entry (probably pointing at your default gateway) or direct to your ISP name servers. nameserver 192.168.1.1 ==== Troubleshoot wireless network problems ==== * [[https://www.blackmoreops.com/2014/09/18/connect-to-wifi-network-from-command-line-in-linux/|connect to wifi network from command-line in linux - from blackmoreops.com]] **Short summary of all the things you need to do in just few lines** root@kali:~# iw dev root@kali:~# ip link set wlan0 up root@kali:~# iw wlan0 scan root@kali:~# wpa_passphrase blackMOREOps >> /etc/wpa_supplicant.conf root@kali:~# wpa_supplicant -i wlan0 -c /etc/wpa_supplicant.conf root@kali:~# iw wlan0 link root@kali:~# dhclient wlan0 root@kali:~# ping 8.8.8.8 (Where wlan0 is wifi adapter and blackMOREOps is SSID) (Add Routing manually) root@kali:~# ip route add default via 10.0.0.138 dev wlan0 ==== To change to static IP address (Redhat/CentOS) ==== As root: vi /etc/sysconfig/network-scripts/ifcfg-eth0 DEVICE=eth0 BOOTPROTO=STATIC IPADDR=192.168.1.111 NETMASK=255.255.255.0 GATEWAY=192.168.1.1 ONBOOT=yes DNS1=8.8.8.8 DNS2=8.8.4.4 Resrart the network interface /etc/init.d/network stop /etc/init.d/network start or service network restart Check name server entry in resolv.conf vi /etc/resolv.conf nameserver 192.168.1.1 ==== Enable processes / services to start at boot time ==== sudo chkconfig httpd on sudo chkconfig mysqld on ==== Run a command on lots of servers in parallel ==== This is a damn fine AIX utility - part of the CSM Distributed Shell. dsh -a "ls -al /etc/apache2/*conf" will list the Apache configuration file on all reachable servers (nodes) ==== Download a gzip file and pipe it into tar ==== cd ~ && wget -O - "https://www.dropbox.com/download?plat=lnx.x86_64" | tar xzf - ==== Check in a script to make sure it is run only by root ==== RUID=`/usr/bin/id|$AWK -F\( '{print $1}'|$AWK -F\= '{print $2}'` if [ ${RUID} != "0" ];then $ECHO "This script must be executed as root" exit 1 fi ==== Set terminal to use Backspace key to erase previous character instead of Control-H ==== Been looking for this for a long time.\\ You can put: stty erase in your .profile but this will be ruined if you do a copy/paste into another file.\\ I wanted a way of doing this without entering the control character in the .profile. Finally stumbled upon it. And it's so simple. Just escape the caret! stty erase \^? Put this in the .profile. It's copy/pastable and it works!\\ If you want CTRL-H to be your erase character, just do this: stty erase \^H ==== Play with the terminal settings and reset them again either side of requesting a password ==== The -g option of stty gives a compact list of all the settings or the terminal and can be used as input to stty OLDCONFIG=`stty -g` # save terminal configuration stty -echo # turn character echoing off echo "Enter password: \c" read PASSWD # get the password stty $OLDCONFIG # restore terminal configuration ==== Reset terminal to "sane" characteristics ==== If you've done a cat of a binary file or something else weird and your terminal is left in a mess, the following key sequence should bring it back to normal stty sane ==== Install OpenOffice on RedHat Enterprise when yum install doesn't! ==== Download Package wget http://sourceforge.net/projects/openofficeorg.mirror/files/4.0.1/binaries/en-US/Apache_OpenOffice_4.0.1_Linux_x86-64_install-rpm_en-US.tar.gz/download -O Apache_OpenOffice_4.0.1_Linux_x86-64_install-rpm_en-US.tar.gz Change to root sudo su - Remove the old stuff yum remove openoffice* libreoffice* Extract Package tar -xvf Apache_OpenOffice_4.0.1* cd en-US Install Package and exit root rpm -Uvh RPMS/*.rpm RPMS/desktop-integration/openoffice4.0-redhat-*.rpm exit Start it openoffice4 ==== What does this do? ==== while IFS= read -r line; do echo "[$(date "+%F %T")] - $line" done < <(iwevent) or while IFS= read -r line; do printf "%s\n%s\n" "$line" "Yohooo! One more package." done < <(tcpdump -i any -nS) ==== List of Special Characters and what they mean ==== From [[https://www.grymoire.com/Unix/Quote.html|Bruce Barnett]] Character Where Meaning csh, sh Execute command # csh, sh, ASCII files Start a comment csh, sh Argument separator ` csh, sh Command substitution " csh, sh Weak Quotes ' csh, sh Strong Quotes \ csh, sh Single Character Quote variable sh, csh Variable variable csh, sh Same as variable | csh, sh Pipe character ^ sh Pipe Character & csh, sh Run program in background ? csh, sh Match one character * csh, sh Match any number of characters ; csh, sh Command separator ;; sh End of Case statement ~ csh Home Directory ~user csh User's Home Directory ! csh History of Commands - Programs Start of optional argument $# csh, sh Number of arguments to script $* csh, sh Arguments to script $@ sh Original arguments to script $- sh Flags passed to shell $? sh Status of previous command $$ sh Process identification number $! sh PID of last background job && sh Short-circuit AND || sh Short-circuit OR . csh, sh Typ. filename extension . sh Source a file and execute as command : sh Nothing command : sh Separates Values in environment variables : csh Variable modifier Character Where Meaning [ ] csh, sh Match range of characters [ ] sh Test %job csh Identifies job Number (cmd;cmd) csh. sh Runs cmd;cmd as a sub-shell { } csh In-line expansions {cmd;cmd } sh Like (cmd;cmd ) without a subshell >ofile csh, sh Standard output >>ofile csh, sh Append to standard output >!file csh Append to file, ignore error if not there >!file csh Output to new file, ignore error if not there >&file csh Send standard & error output to file <&digit sh Switch Standard Input to file <&- sh Close Standard Input >&digit sh Switch Standard Output to file >&- sh Close Standard Output digit1<&digit2 sh Connect digit2 to digit1 digit<&- sh Close file digit digit2>&digit1 sh Connect digit2 to digit1 digit>&- sh Close file digit