on an ASG320 cluster with 1 GB each (active-passive) I experienced massive RAM und swap usage on both machines:
Master: RAM 70%, SWAP 99%
Slave: RAM 81%, SWAP 93%
Approx. 40% of the memory is consumed by tow slon processes. In the system log I find theses messages all the time:
2009:07:10-09:00:01 (none) /usr/sbin/cron[17052]: (root) CMD ( /usr/local/bin/reportcontrol.sh)
2009:07:10-09:00:02 (none) postgres[17062]: [3-1] ERROR: relation "pfilter" does not exist
2009:07:10-09:00:02 (none) postgres[17062]: [3-2] STATEMENT: INSERT INTO pfilter (dayabs, srcip, srccountry, dstip, dstcountry, svc, packets) VALUES ($1,$2,$3,$4,$5,$6,$7)
2009:07:10-09:00:02 (none) postgres[17062]: [4-1] ERROR: prepared statement "dbdpg_1" does not exist
[...]
2009:07:10-09:00:02 (none) ulogd[4684]: pg1: error allocating keyset: Success
[...]
2009:07:10-09:17:01 (none) /usr/sbin/cron[18485]: (root) CMD ( nice -n19 /usr/local/bin/gen_inline_reporting_data.plx)
2009:07:10-09:17:02 (none) ulogd[4684]: pg1: error allocating keyset: Success
2009:07:10-09:17:02 (none) postgres[18486]: [3-1] ERROR: relation "auth" does not exist
2009:07:10-09:17:02 (none) postgres[18486]: [3-2] STATEMENT: SELECT dayabs,secabs,int2ip(srcip),username,authresult,facility FROM auth WHERE facility LIKE 'webadmin' AND dayabs
2009:07:10-09:17:02 (none) postgres[18486]: [3-3] = 733598 ORDER BY secabs DESC
2009:07:10-09:17:02 (none) postgres[18486]: [4-1] ERROR: relation "auth" does not exist
2009:07:10-09:17:02 (none) postgres[18486]: [4-2] STATEMENT: SELECT dayabs,secabs,int2ip(srcip),username,authresult,facility FROM auth WHERE authresult LIKE 'failed' AND dayabs
2009:07:10-09:17:02 (none) postgres[18486]: [4-3] = 733598 ORDER BY secabs DESC
As CPU usage is quite low (around 10%) and only the packet filter and vpn functions are used there isn't a performance problem at present. Do I have to worry about this issue?
Thanks & Greetings
marc
This thread was automatically locked due to age.