Skip to content

Latest commit

 

History

History
139 lines (95 loc) · 4.76 KB

README.md

File metadata and controls

139 lines (95 loc) · 4.76 KB

quick-log

Tools for logging program execution outputs or login sessions, mostly bash functions and aliases, goes in your .bashrc

TL;DR

Yesterday I wrote a set of tools for logging program runs and sessions. Can't believe I didn't write them 20 years ago.... Simple usage example:

Runs the program barman and creates ~/logs/barman-2019-12-05T23:54:43 and logs all the program output to the log file.

log-plain barman -v check-backup mtv-pg-01 --run 20191127T045147

View the newly created log file in less

log-less barman

Install

Just add it to your .bashrc, or source it from your .bashrc.

The ugly way (not really recommended):

cat quick-log.bash >> ~/.bashrc

The clean way:

echo "source /home/wwalker/git/quick-log/quick-log.bash" >> ~/.bashrc

Why?

As a sysadmin, I'm always needing to log a login session, or the output of some command, which may be a binary. To solve these problems, I've written a few things.

For recording a login session, the easiest is to use the program script. But, I found that calling script (without any arguments) would often result in me overwriting the output ( the file typescript, which script creates by default) from a previous session. So I created an scr alias. All it does is prevent overwriting. You need a ~/tmp directory.

$ alias scr='script ~/tmp/typescript-$(date +%FT%T)'
$ scr
Script started, file is /home/wwalker/tmp/typescript-2019-12-05T18:56:31
$

That is great when I want to catch an interactive session.

However, when I want to log the output of a command (script or binary), I either want the exact output, or I want the output with timestamps in front of each line. So I wrote these some bash functions:

  • log-plain <command and its arguments> - run the command passed in as arguments, automatically writing stdout and stderr to a new unique log file ( ~/logs/-YYYY-MM-DD-HH:MM:SS )
  • log-time <command and its arguments> - just like log-plain, except it prefixes everyline with a timestamp ( YYYY-MM-DD-HH:MM:SS.micros )
  • log-newest <program-name> - returns the name the newest log file for program-name
  • log-less <program-name> - opens the newest log file for program-name using less
  • log-view <program-name> - opens the newest log file for program-name using view
  • log-grep <program-name> <grep arguments and regular expression> - runs grep on the newest log file for program-name

Just run your command like you normally would prefixing it with either log-plain or log-time:

log-plain barman -v check-backup mtv-pg-01 --run 20191127T045147

or

log-time barman -v check-backup mtv-pg-01 --run 20191127T045147

This will painlessly create 2 files like this:

wwalker@polonium:~ ✓ $ ls -l ~/logs/barman*
-rw-r--r-- 1 wwalker wwalker  7623 2019-12-05 18:25:07.873 /home/wwalker/logs/barman-2019-12-05T18:25:06
-rw-r--r-- 1 wwalker wwalker 10296 2019-12-05 18:34:59.546 /home/wwalker/logs/barman-2019-12-05T18:34:57

But, Wait there's more!!

I didn't want to have to do an ls to find the name of the log file that log-plain or log-time just created for me.

So, you run your command with log-plain (or log-time), and then you just use log-less or log-view (I'll probably create an emacs log for Those people):

log-plain barman -v check-backup mtv-pg-01 --run 20191127T045147
log-less barman

That's it, no ls -lrt ~/tmp, no tab completion games to find the file name. Just run log-less (or log-view if you like using vim to look at logs).

But, Wait there's more!!

log-less, log-view, and log-newest only look at the first argument; so, you can be lazy:

log-plain make -D prefix=/opt

Now, you just up arrow and change "just" to "less", and your looking at the recent log file:

log-less make -D prefix=/opt

log-less (and log-view and log-newest) just ignore the extra arguments.

But, wait! There's more!

You get programmable completion (tab completion) for whatever commands you put after log-plain and log-time.

But, wait! There's even more!

"I use grep all the time on my log files" - And the answer is, you guessed it, log-grep

First, get the text from all 800 servers' /etc/cron.d/atop files that are broken:

log-plain fordcrun 'grep -F "0 0 * * root" /etc/cron.d/atop'

Then get the hostnames (on the line above the output in the file :-) ) with log-grep (without having to look up the log file name):

wwalker@polonium:~ ✓ $ log-grep fordcrun  -B1 -F "0 0 * * root"
int-salt-01:
    0 0 * * root systemctl restart atop
--
rcn-pg-01:
    0 0 * * root systemctl restart atop
rcn-pg-02:
    0 0 * * root systemctl restart atop
rcn-pg-03:
    0 0 * * root systemctl restart atop

Coming soon to a repo near you:

  • command completion of the log file names for when you need to look at an older log.