Automating MustGather and other diagnostic collection using AutoGather

The manual collection of Performance MustGather is quite a complicated series of steps - see for example the DOORS Next Performance MustGather.

AutoGather is a new diagnostic application to help by automating large parts of the MustGather collection, which can also be used for other diagnostic collection needs.

Introduction

The manual collection of Performance MustGather is quite a complicated series of steps - see for example the DOORS Next Performance MustGather - often requiring an IBM support engineer to run a call coordinating actions among several client personnel including at least one JazzAdmin and a user; the user reproduces the problem while the other client personnel capture information, such as triggering javacores every thirty seconds during the reproduction. Around this main collection there are various pages and files which need to be combined with the ISADC logs to attach to the support case. This complexity often causes some of the files to be missed and this typically won't be discovered until after the call has finished, and may require the complete collection to be repeated.

AutoGather is a new diagnostic application to help by automating large parts of the MustGather collection - requiring only a JazzAdmin and the user to reproduce the problem - and speeds up the collection, with no possibility that parts of the collection will be missed.

What AutoGather can do

AutoGather is a Java application which uses available APIs of the running server to collect more than just the log files. A particular difference from ISADC is that it can trigger and collect javacores at controlled intervals - this is valuable for analysis of many issues including performance problems.

In addition to expediting collection of Performance MustGather, AutoGather can meet other needs for logs collection, such as:

• Remote Gather - simplifying log collection from multiple servers
• "Super" ISADC - collects everything that ISADC collects and also collects javacores, verbosegc, server startup and configuration details
• Monitoring - run AutoGather for a period (e.g. one hour, or twenty four hours) collecting javacores and other information at regular intervals
• Gathering DEBUG loggers - automatically enables loggers to DEBUG at the start of the collection and reverts these when the collection finished

Collection modes for AutoGather

AutoGather can be used in three modes of operation:

• Normal: this mode is intended to directly address the collection of Performance MustGather diagnostics and logs. AutoGather must be installed on the application server and is run there, collecting files and logging in to the ELM server to gather other diagnostic information - however if the ELM deployment is using a web-based authentication scheme such as SAML or ADFS this mode can't be used.
• Remote: this mode is intended for use when
• Anonymous: this mode doesn't require a username and password to log into ELM - in that respect it's most like ISADC. However the collection is more limited, so this isn't suitable for the full Performance MustGather collection. However this mode also requires no/minimal configuration as it collects information through the filesystem. Because no login is required this can collect disagnostics when the ELM application is using a web-based authentication scheme like SAML or ADFS, whoich isn't possible in the other two modes.

What is collected in each mode

Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text

Heading 2 (use sentence-style capitalization)

Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text Sub-Section text

Heading 2 (use sentence-style capitalization)

Heading 1

Related topics: Deployment web home, Deployment web home

External links:

• IBM

Additional contributors: TWikiUser, TWikiUser