Hi all,
I have a zabbix server (4.0.8), 3 active proxies(4.0.8) and a bunch of agents (roughly 1500, 2.2).
I wanted to get a high NVPS load on a proxy, so I reduced the frequency of my polling to 20 seconds, and had 1000 of them monitored by one proxy.
The web GUI tells my I am over 3700 NVPS on the server:
but when I check the data, I don't have a polling every 20 seconds but rather between 60 and 90 seconds.:
Also I have a huge queue, but not so many measurements that are delayed over 5 minutes (although more than 7000 delayed by more than 1 minute is quite a drawback):
Migrating the agents right away is not an option as I have a grand total of over 20,000 nodes and it is planned for next year.
So far, proxy measurements don't seem off (values for the loaded proxy are in green):
Can you help me find out why so many values are discarded ?
I have a zabbix server (4.0.8), 3 active proxies(4.0.8) and a bunch of agents (roughly 1500, 2.2).
I wanted to get a high NVPS load on a proxy, so I reduced the frequency of my polling to 20 seconds, and had 1000 of them monitored by one proxy.
The web GUI tells my I am over 3700 NVPS on the server:
but when I check the data, I don't have a polling every 20 seconds but rather between 60 and 90 seconds.:
Also I have a huge queue, but not so many measurements that are delayed over 5 minutes (although more than 7000 delayed by more than 1 minute is quite a drawback):
Migrating the agents right away is not an option as I have a grand total of over 20,000 nodes and it is planned for next year.
So far, proxy measurements don't seem off (values for the loaded proxy are in green):
Can you help me find out why so many values are discarded ?
Comment