Hello all,
My name is max and I am new to this forum. I am a beginner to Bash Scripting and been having a real problem trying to compare two log files. Long story short...
I have a known good wrapper.log file and I would like to have other new Wrapper.log files to compare next to it to see if any services did not start properly.
Normally I could just use the diff command or the cmp command in a script but this log file does not necessarily boot in the same lined order every time. So I think the only thing I am left with is using bash arrays? Could anyone point me in the right direction on how to start this project? MUCH MUCH thanks in advance.

This is a little snippit of the known good wrapper.log file below

STATUS | wrapper | 2011/09/28 19:45:00 | --> Wrapper Started as Daemon
STATUS | wrapper | 2011/09/28 19:45:00 | Java Service Wrapper Standard Edition 64-bit 3.4.1
STATUS | wrapper | 2011/09/28 19:45:00 | Copyright (C) 1999-2010 Tanuki Software, Ltd. All Rights Reserved.
STATUS | wrapper | 2011/09/28 19:45:00 | http://wrapper.tanukisoftware.org
STATUS | wrapper | 2011/09/28 19:45:00 | Licensed to TradeCard, Inc. for TradeCard Linux 64
STATUS | wrapper | 2011/09/28 19:45:00 |
INFO | wrapperp | 2011/09/28 19:45:00 | port 32000 already in use, using port 32001 instead.
STATUS | wrapper | 2011/09/28 19:45:00 | Launching a JVM...
INFO | jvm 1 | 2011/09/28 19:45:00 | Please use CMSClassUnloadingEnabled in place of CMSPermGenSweepingEnabled in the future
INFO | jvm 1 | 2011/09/28 19:45:01 | java version "1.6.0_22"
INFO | jvm 1 | 2011/09/28 19:45:01 | Java(TM) SE Runtime Environment (build 1.6.0_22-b04)
INFO | jvm 1 | 2011/09/28 19:45:01 | Java HotSpot(TM) 64-Bit Server VM (build 17.1-b03, mixed mode)

So, you need to first order the lines (e.g. using service name as the field to be used for ordering), and then 'diff' the ordered lines - say, saving them first in files.

If you no need orginal order of lines in theses log files, maybe use "sort" to obtain differences. Also, for later looking these lines in orginal log, you can add column with orginal line number, for ex. "grep" can do it.

I, too, am looking for a solution to the problem of comparing log files. The previous replies have focused on the problem that the lines may not always be in the same order. However, the main problem is that the time stamps are always different. Just being able to execute a "diff" while ignoring the different time stamps would help. Does anyone have any suggestions to achieve that?

Write a parser, then compare, using a custom function, the resulting from parser data structures of the two files to be compared. The custom function can ignore, for example, timestamps.

The proposed solution is ages old and is widely used.