I have a Zabbix server and four proxies. Housekeeping on three of the proxies runs in lockstep with the server's housekeeping, as shown by Zabbix graphs. By "lockstep" I mean that the housekeeping starts at the same time and runs for the same duration. One of the proxy's housekeeping is not in lockstep with any other box.
All five boxes (the one server and the four proxies) are using independent, local postgresql servers (DBHost=localhost).
I discovered this apparent synchronization after I switched one of my proxies from spinning disk to SSD in order to find out how that affected the speed of housekeeping.
What might cause the housekeeping of most of my Zabbix boxes to appear to run in lockstep? Is this real, or are the graphs likely to be erroneous?
Versions:
All five boxes (the one server and the four proxies) are using independent, local postgresql servers (DBHost=localhost).
I discovered this apparent synchronization after I switched one of my proxies from spinning disk to SSD in order to find out how that affected the speed of housekeeping.
What might cause the housekeeping of most of my Zabbix boxes to appear to run in lockstep? Is this real, or are the graphs likely to be erroneous?
Versions:
- Zabbix: 5.0.8
- OS: Debian 11.2 and 11.3
- Postgresql: 13.5-0+deb11u1
Attached Files