Agent.ping not scheduling after going KO

[Erwan A.]

Hi everyone,

I'm going to take control of a Zabbix server in my new job and i found a weird behavior.
Zabbix server version : 3.4 (docker)
Zabbix agent version : 3.4.7
Servers OS : Ubuntu
Sometimes, when zabbix agent is unreachable, zabbix stop check it's state.
For example, this morning à 11h50 we have lost one of our celery's server :

Problem started at 11:50:00 on 2019.06.18 Problem name: Zabbix agent on Backend SaaS Celery is unreachable for 5 minutes

We found and solved the problem but, since 11h50, zabbix never renew agent.ping status however other metrics are OK.
If I check agent.ping from zabbix server, it's ok :

But as you could see, others checks are OK but agent.ping didn't re-check since 11h50 :


Could you help me to find my problem ?
Thanks.

Edit : Problem solved. Server was swapping a lot, swap at 100%. I had to restart celeryd and know it's OK.

[1berto]

Am i wrong or the Update interval of the ping is lower than the other items?
If it's, that's the cause.