VMWARE "Timeout was reached"

[rockandstone]

Hello!
I am using good template for VMware Hypervisor but after 3-4 months I have a problem: "Timeout was reached" in all my vcenter's item.
We updated Virtual Center to V.7.0.2 Update 2 - Build release 17694817 and then problem starts to appear.

I edited settings in zabbix_proxy.conf three months ago:

### Option: StartVMwareCollectors
StartVMwareCollectors=5
VMwareFrequency=60
VMwarePerfFrequency=60
VMwareCacheSize=256M
VMwareTimeout=40

In zabbix_server.log I found continuous errors (every 10-20min):

"Numbers here":item "host:item_key[...]" became not supported: Timeout was reached
"Numbers here":item "host:item_key[...]" became not supported: became supported
"Numbers here" error reason for "host:item_key[...]" changed: Performance counter data is not available.

I hope that someone has encountered this kind of problem before ...

Thank you for your help

[gofree]

hi

try to increase timeout ( up to 300s ), see last line on link bellow

https://www.zabbix.com/documentation/current/manual/appendix/config/zabbix_proxy

[rockandstone]

Ok I appreciate your idea and I think it will work, but would it be a "polished" workaround? I mean... 300 second of timeout will be too much since every minute I get data?

[rockandstone]

Ok I appreciate your idea and I think it will work, but would it be a "polished" workaround? I mean... 300 second of timeout will be too much since every minute I get data?

Sorry I replied in a bad way, I am new in zabbix forum ^^

[gofree]

Ok I appreciate your idea and I think it will work, but would it be a "polished" workaround? I mean... 300 second of timeout will be too much since every minute I get data?

up to 300s means it can be 300 max ( it can be also 59 )

P.S. 1 min interval could the cause ( not saying it is ) that it timeouts - if you have plenty VM it can cause it ( personally I consider interval 1min with zabbix in large quantities a road to hell )

[rockandstone]

Ok thank you for your explanation...
I wonder if I should rise update interval value, even if I need them every minute (cpu usage, latency, ram, disk usage, ...). You called 1m interval a road to hell and this scared me, because we have 35+vm monitored. I will update you with fresh news asap!

[gofree]

I mean its road to hell if you have thousands items with 1 minute interval without measures like throttling, db partitioning, preprocessing.

[rockandstone]

Thank you for your help!

I have edited proxy config file:
StartVMwareCollectors=5
VMwareFrequency=60
VMwarePerfFrequency=60
VMwareCacheSize=256M
VMwareTimeout=60

And it worked like a charm... I increased refresh rate too (from 1m to 5m), but I have to be careful because of housekeeper...