Hello,
I've run into an situation where it appears the only way out is for me to manually set all pending rows (status=0) to failed (status=2). My question is if I do this via mysql query do I need to restart my zabbix-server instance. I would prefer not to, as it currently takes ~1 hour to complete a full stop and start of the service.
I had to restart the Zabbix-server process about a week ago and the alerter process has been pegged at 100% since then. Looking at the alerts table I see that there are ~2.7 million rows that match status=0 and alerttype=0.
I don't need the actions from the week to actually complete at this point, but I would like to get back to seeing the emails and other actions happen on time.
To provide some context on my zabbix installation, we are currently running 275+ proxies, average of 10kvps, with ~4.7 million items and ~3.5 million triggers. We are running zabbix 3.0.29 on an ubuntu 16.04 host.
If there is another way to handle this I am open to suggestions.
I've run into an situation where it appears the only way out is for me to manually set all pending rows (status=0) to failed (status=2). My question is if I do this via mysql query do I need to restart my zabbix-server instance. I would prefer not to, as it currently takes ~1 hour to complete a full stop and start of the service.
I had to restart the Zabbix-server process about a week ago and the alerter process has been pegged at 100% since then. Looking at the alerts table I see that there are ~2.7 million rows that match status=0 and alerttype=0.
I don't need the actions from the week to actually complete at this point, but I would like to get back to seeing the emails and other actions happen on time.
To provide some context on my zabbix installation, we are currently running 275+ proxies, average of 10kvps, with ~4.7 million items and ~3.5 million triggers. We are running zabbix 3.0.29 on an ubuntu 16.04 host.
If there is another way to handle this I am open to suggestions.
Comment