My apologies for my previous attempt. -------
UV 10.0.8, Solaris 8... On Thursday and Friday night we've had our system grind to a halt with 'Available memory exceeded' messages... At first blush I would think I should have to increase max shared memory (shminfo_shmmax for Solaris), but I've got a fairly large value all ready (536870912, or 512mb). I am seeing errors in /var/adm/messages that coincide with the failures. However they warn of swap space deficiencies... for example: Jul 12 22:43:55 ccses genunix: [ID 470503 kern.warning] WARNING: Sorry, no swap space to grow stack for pid 22440 (uvsh) Jul 12 22:43:55 ccses last message repeated 9 times I found the &PH& entry for a job that had the above PID. The job up and died with out giving any error messages. The job is not new, and had been running every night without incident. I plan on having our normal nightly processing run tonight, with the exception of the one suspect job. Any thoughts or suggestions? Mark Hennessey State of Connecticut Department of Social Services Information Technology Services Child Support Systems Voice: 860-424-5261 Fax: 860-424-4813 ------- u2-users mailing list [email protected] To unsubscribe please visit http://listserver.u2ug.org/
