Translating with DrWatson… this can take a few seconds the first time.
This is a genuine, complex translation. DrWatson protects commands, error codes, and log output while naturally translating the surrounding text. It’s translated once and saved.
Rajat Agrawal asked what causes a server to keep adding virtual shared memory segments. SangGyu Jeong suggested unclosed statements and checking sessions with onstat -g ses. Mark Jalkiewicz cited defect IT19869; Art Kagel named big queries and many sessions. Doug Lawry advised onstat -g mem sorted for virtual pools, and noted APAR IT31418 (RA_req_* pool leak with auto update statistics) fixed in 12.10.xC14 and 14.10.xC4.
Auto-generated by Claude from the posts below — may be imperfect; read the full thread.
Rajat Agrawal — source: IBM Community (ConnectedCommunity.org) Informix forum
Hi,
I have a server which keeps on adding virtual shared memory segments (SHMADD).
Want to check if there is any culprit process contributing to this addition or I really need to increase SHMVIRTSIZE.
Possible causes I know might be:
1. Increase in number of connections.
2. Dynamically allocated locks.
3. Backup.
What else can I check to see if there is any Application process which can be tuned ?
What are the contributors for virtual shared memory segment addition ?
Thanks for the help !!
------------------------------
Rajat Agrawal
------------------------------
#Informix
↪ replying to Rajat Agrawal
SangGyu Jeong — source: IBM Community (ConnectedCommunity.org) Informix forum
Hi Rajat,
In rare cases, a memory leak may occur because the statement does not close in the application code.
There may be sessions that increase memory usage with poorly written code that way.
You may also want to monitor high session memory usage with the onstat -g ses command or sysmaster information.
------------------------------
SangGyu Jeong
Software Engineer
Infrasoft
Seoul Korea, Republic of
------------------------------
↪ replying to Rajat Agrawal
Mark Jalkiewicz — source: IBM Community (ConnectedCommunity.org) Informix forum
Art Kagel — source: IBM Community (ConnectedCommunity.org) Informix forum
The other main cause is big complex queries and large numbers of concurrently active sessions.
Art
↪ replying to SangGyu Jeong
Rajat Agrawal — source: IBM Community (ConnectedCommunity.org) Informix forum
Thanks SangGyu !
------------------------------
Rajat Agrawal
------------------------------
↪ replying to Rajat Agrawal
Doug Lawry — source: IBM Community (ConnectedCommunity.org) Informix forum
Have a look at the output of "onstat -g mem". The largest pools can be viewed with:
onstat -g mem | sort -k4 -nr | head -20That might give some clues. You are only interested in those with class V (virtual) as others are fixed.
If this is a gradual long-term problem, you have a very large number of pools named RA_req_*, and Automated Update Statistics is enabled in the Task Scheduler, you may have encountered a bug we worked through with IBM who reported this:
"T
he defect APAR IT31418 will be fixed in the next maintenance release of IDS 12.10 and 14.10:
idsdb00104976 -> IDS 12.10.xC14 --> planning to be released around 12 March
idsdb00105034 -> IDS 14.10.xC4 --> planning to be released around 11 June
This defect has been caused by a memory leak that seems to happen with RA_REQ_PAGELIST request only."
------------------------------
Doug Lawry
------------------------------
↪ replying to Doug Lawry
SangGyu Jeong — source: IBM Community (ConnectedCommunity.org) Informix forum
Hi Doug,
Thank you for sharing information about the defects and release date of the next fix pack!
------------------------------
SangGyu Jeong
Software Engineer
Infrasoft
Seoul Korea, Republic of
------------------------------
We use strictly necessary cookies to make this site work. With your
consent we’d also use optional cookies for analytics and marketing. You can accept all,
reject all, or choose. Read our Cookie Policy.