Zabbix 1.8.1 queue falls far behind -- worked well on 1.8
Yesterday I updated a production Zabbix 1.8 box running on a 64-bit Debian box to 1.8.1. Prior to that, Zabbix had been on top of the queues without any problems. Now, it starts out with a rather large queue and just when it's about got it cleared up, it snowballs into an untenable backlog and essentially stops monitoring. Nothing interesting in the logs of either Zabbix or MySQL, and the CPU and memory load averages look fine. MySQL queries complete quickly, with about 4GB dedicated to SQL -- primarily to InnoDB.
For reference, it's running under a Xen hypervisor, but the Zabbix server has 4 CPU cores, 8GB RAM, and a dedicated RAID array allocated to it.
Number of hosts: 144
Number of items: 8423
Number of triggers: 4056
New values per second: 282.0927
Any suggestions?
For reference, it's running under a Xen hypervisor, but the Zabbix server has 4 CPU cores, 8GB RAM, and a dedicated RAID array allocated to it.
Number of hosts: 144
Number of items: 8423
Number of triggers: 4056
New values per second: 282.0927
Any suggestions?
Comment