My system is Zabbix 1.8.10 and Postgres 8.0 in a /var/log/messages is logging many postgress error messages follows:
postgres[28915]: [140-198] #011update items set lastclock=1361938976,,prevorgvalue='32345155',prev value=lastvalue,lastvalue='0' where itemid=37142;
(not the same itemid every time ) The system is logging so many of these that the log msg rate limits them. Its usually repeating about every 60 seconds.
I do notice with the admintrator ->queue menu on zabbix dashboard, that the queueing has a few requests queued for over an hour.
A google search on the postgress forum turned up this http://www.postgresql.org/message-id...mail.gmail.com Essentially Jeff Janes suggested the error messages may be a symptom of deadlock: He states " The problem you are seeing is with deadlocks, not with sharelocks.(The fact that transaction waits are implemented as ShareLocks is an internal implementation, and perhaps the error message should be
changed so as not to leak that to the user).The problem is not with contention in general, but just the special type of contention that leads to cyclic dependencies, i.e. deadlocks. He also suggests "One possibility is that they are
bunching unrelated real-world transactions into fewer database
transactions in an attempt to improve "group commit" performance. In
which case, they should probably not do that and instead turn
synchronous commit off." He goes on to indicate if there “If Zabbix retries the transaction when it gets thrown a deadlock
error, then there is not really a problem, just a cluttering of your
log file. Unless they are happening extremely frequently and creating
performance problems.”
My required server performance is only 130 and number of items being monitored about 7000se . Dell R710 servers should be able handle it
I also tried doubling my DBsyncers to 8..but that didn’t help, nor did reducing them to two. I did receive checkpoint hint messages in the log also, however increasing the segments from a default of 3 to ten helped alleviate that error message.
I set the housekeeping disable to 1 but I still received the errors. My trends and history are set to days for all items..
If I disable a few of the monitored hosts in zabbix, then the Required server performance figure goes down to 112 similar to the OCA system. And deadlocks are greatly alleviated.
What can be happening.
Thanks
John Matchett
postgres[28915]: [140-198] #011update items set lastclock=1361938976,,prevorgvalue='32345155',prev value=lastvalue,lastvalue='0' where itemid=37142;
(not the same itemid every time ) The system is logging so many of these that the log msg rate limits them. Its usually repeating about every 60 seconds.
I do notice with the admintrator ->queue menu on zabbix dashboard, that the queueing has a few requests queued for over an hour.
A google search on the postgress forum turned up this http://www.postgresql.org/message-id...mail.gmail.com Essentially Jeff Janes suggested the error messages may be a symptom of deadlock: He states " The problem you are seeing is with deadlocks, not with sharelocks.(The fact that transaction waits are implemented as ShareLocks is an internal implementation, and perhaps the error message should be
changed so as not to leak that to the user).The problem is not with contention in general, but just the special type of contention that leads to cyclic dependencies, i.e. deadlocks. He also suggests "One possibility is that they are
bunching unrelated real-world transactions into fewer database
transactions in an attempt to improve "group commit" performance. In
which case, they should probably not do that and instead turn
synchronous commit off." He goes on to indicate if there “If Zabbix retries the transaction when it gets thrown a deadlock
error, then there is not really a problem, just a cluttering of your
log file. Unless they are happening extremely frequently and creating
performance problems.”
My required server performance is only 130 and number of items being monitored about 7000se . Dell R710 servers should be able handle it
I also tried doubling my DBsyncers to 8..but that didn’t help, nor did reducing them to two. I did receive checkpoint hint messages in the log also, however increasing the segments from a default of 3 to ten helped alleviate that error message.
I set the housekeeping disable to 1 but I still received the errors. My trends and history are set to days for all items..
If I disable a few of the monitored hosts in zabbix, then the Required server performance figure goes down to 112 similar to the OCA system. And deadlocks are greatly alleviated.
What can be happening.
Thanks
John Matchett