Performance data from topasout <xmwlm recording>

Starting from AIX 5.3 TL 04 an automatic collection of performance information is available for a single AIX image. The xmwlm process is in charge of data collection and it is automatically started at boot time due to the following line in /etc/inittab:

xmdaily:2:once:/usr/bin/xmwlm -L 2>&1 >/dev/null # Start xmwlm daily recording

Every day the process stores data into a binary file named xmwlm.<YYMMDD> in /etc/perf_daily. A new file is created every day. In order to limit disk space usage, only current and previous day data are retained and older files are deleted. There are plans to extend this data retention period, but you should check and copy data files if you need to store them for a longer period.

The topasout command is used to extract data from the binary file and format the output in several ways. In order to use pGraph, just use the following syntax:

topasout <binary file>

The command will create a new text file in the same directory of the binary file with the name <binary file>_01. This file can be read by pGraph either in interactive-mode and in batch mode.

Single file

When pGraph reads a single xmwlm file, it provides the GUI shown by the following figure:

The same set of information is provided in batch mode.

Multiple files from the same host

pGraph is capable to manage multiple files created by xmlm as if they were a single file. Put all files that require analysis in the same directory and use pGraph to read the directory with the menu or flag that states single host, multiple files. pGraph will internally concatenate the data and handle the set of files exactly as if a single file was provided. There is no requirement on the file names.

Multiple files from multiple hosts

It is common to analyze multiple systems at a time, especially when LPARs are involved. In this case pGraph focuses on CPU performance data related to multiple hosts where each file represents a single host.

pGraph extracts CPU data from each file and creates the following performance graphs:

• Average CPU statistics (User, System, Wait, Idle) for each host.
• Physical CPU usage for each host, if it is a micro partition.
• The sum of physical CPU usage of all micro partitions, if present.