Greetings list... I'm running RH 7.3 with the latest kernel update(2.4.20-20.7). The machine is a MySQL server that gets most of its activity overnight during some batch processing. Since upgrading from the original kernel distributed with 7.3, we've seen swap space used when none was used before. I've tried tuning the MySQL memory usage, but that doesn't affect the swap usage. The server is also significantly slower now, taking about twice the amount of time for an evening's updates.
The server has Dual processors, 4GB RAM and a large SCSI array setup as an ext3 filesystem. The the largest & most active datbase file is around 8GB. When typically running, free reports over 3GB used in cache, 200-400MB swapped and low free memory. The MySQL process shows about 500MB ussage. In researching this problem I found Redhat Bug #89226 (https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=89226). Wow! That looks like my problem. The 2.4.20-20.7 kernel release has a changelog that references updating -rmap to correct this problem, but the bug is not reported as fixed/closed. I haven't seen any new discussion of the bug since early Aug. Does anyone have any insight/explanation or possible work-arounds for this problem? I've tried tweaking the bdflush settings as discussed in the bug, but I'm not seeing any improvement. Has this problem ~really~ been fixed in 2.4.20-20.7? Is there something else I need to change/update on my server? Any work-arounds for this problem? TIA for any help...I've run out of things to try to correct the problem, short of going back the the kernel with the SMP bug! Regards, tom. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
