Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.
Comment: Migrated to Confluence 5.3

...

So as a starter lets enable verbose:gc on the tests runs and monitor the CPU usage additional monitoring can be added to the suite but the first step is to have the ability to gather the data.

Table of Contents
maxLevel5
minlevel3

Data collection

Current goals are to focus on the Java broker by collecting CPU usage (measured via top -p), and GC data as written by the JVM with verbose gc logging. This same information could be gathered for the client JVM however it would require modification to the existing scripts to log the data.

...

As the log entries are timestamped in seconds since VM startup if we want to correlate these values with the CPU or client log then we must convert them to real times. This can be done by looking at the 'Access' time of the gc.log. When retriving the log file (or after starting the JVM) we can gather this information using 'stat':

No Format
title"Output from stat"
  File: `2009-05-22-1016/broker-results/logging/gc.log'
  Size: 1183063         Blocks: 2320       IO Block: 4096   regular file
Device: fd00h/64768d    Inode: 363245      Links: 1
Access: (0644/-rw-r--r--)  Uid: (  500/ritchiem)   Gid: (  500/ritchiem)
Access: 2009-05-29 11:24:00.000000000 +0100
Modify: 2009-05-22 13:42:22.000000000 +0100
Change: 2009-05-22 13:42:22.000000000 +0100

...

The results of the GC and CPU can be put through gnuplot and graphed here is an example of what it might look like: