Question & Answer
Question
MustGather information for 100% CPU usage on Linux. Gathering this information before calling IBM® support helps familiarize you with the troubleshooting process and saves you time.
Answer
- Collecting data for diagnosing 100% CPU usage on versions 11.70.xC1 and later
- Collecting information for versions prior to 11.70.xC1
- Submitting information to IBM Support
In versions of the Informix Server engine 11.70.xC1 and later ifxcollect can be used to gather diagnostic information. To gather information for 100% CPU usage execute the following command while experiencing problems:
$INFORMIXDIR/bin/ifxcollect -c performance -s cpu
Executing ifxcollect will create a tar file in $INFORMIXDIR/isa/data that can be uploaded to IBM for review. The format of the file is ifxcollect.<timestamp>.tar. Uploading the file can be accomplished using any of the methods described in the "Submitting information to IBM Support section".
For additional information on ifxcollect see the following link:
Collecting data with the ifxcollect tool
For all versions of Informix Server prior to 11.70.xC1, gather the following information for any problem:
MustGather: General Information to collect for Informix Family product problems
For all versions of Informix Server prior to 11.70.xC1 gather the following information while the CPU is at 100% utilization:
Read the Linux man pages and IDS documentation for usage and other information about the following commands, if necessary. Run the following commands and collect the output:
- sar -u -o sarfile 5 60 (if installed and available for usage)
- sar -Af sarfile >sar_normal.out
- vmstat 5 60 > vmstat.out
- iostat 5 60 > iostat.out
- for i in 1 2 3 4 5; do ps aux -A >>ps.out; sleep 60; done
- onstat -a > onstat_a.out
- onstat -c > config.out
- for i in 1 2 3 4 5; onstat -g stk all >> onstat_stk.out; sleep 60; done
- for i in 1 2 3 4 5; onstat -g ath >> onstat_ath.out; sleep 60; done
- for i in 1 2 3 4 5; onstat -g rea >> onstat.rea.out; sleep 60; done
- for i in 1 2 3 4 5; onstat -g act >> onstat.act.out; sleep 60; done
- for i in 1 2 3 4 5; onstat -D >>onstat.D.out; sleep 60; done
- for i in 1 2 3 4 5; onstat -g iof >> onstat.iof.out; sleep 60; done
Additionally, this is a list files to collect which are normally (or optimally) generated for this problem:
- online.log
- Any core files or AF assertion failure files generated during the problem.
Once you have collected your information, you can begin Problem Determination through the product Support web page, or simply submit the diagnostic information to IBM support.
You can submit files using one of following methods to help speed problem diagnosis:
- IBM Support Assistant (ISA)
- Electronic Service Request (ESR)
- FTP to the Enhanced Customer Data Repository (ECuRep)
See the Related information article, Submitting diagnostic information to IBM Technical Support for problem determination
Related Information
[{"Product":{"code":"SSGU8G","label":"Informix Servers"},"Business Unit":{"code":"BU053","label":"Cloud & Data Platform"},"Component":"--","Platform":[{"code":"PF016","label":"Linux"}],"Version":"10.0;11.1;11.5;7.3;9.4;11.7","Edition":"","Line of Business":{"code":"","label":""}}]
Was this topic helpful?
Document Information
Modified date:
16 June 2018
UID
swg21271764