Hello, Our Red Hat Enterprise Linux ES 2.1 (Intel) servers are experiencing problems with system using little by little all the free semaphore arrays (128 propably is the limit?), after which for example Apache won't be able to restart anymore. The semaphores then need to be deleted manually, otherwise they seem to stay "reserved" forever, even if the apache processes are killed/restarted. The problem seems to be similar than with RH Linux 7.x -systems, as described in the RH errata RHBA-2002:273, where a bug in the "mm" package was fixed. If I remember correctly, the Enterprise 2.1 -product line is based on RH Linux 7.2, so the problem might well be the same. I couldn't find any errata/bugfixes on this for the Enterprise ES 2.1 server on red hat website/rhn. Has anyone else come across this same problem, any ideas for a workaround? Or does anyone know whether Red Hat is planning to build a bug fix for this for their Enterprise 2.1 series? Many thanks for any advice, Ari Kuukka -- redhat-list mailing list unsubscribe mailto:redhat-list-request@xxxxxxxxxx?subject=unsubscribe https://www.redhat.com/mailman/listinfo/redhat-list
