Zabbix no longer monitoring after upgrade to 1.8.4

[Missbook]

After upgrading from 1.8.3 to 1.8.4 this morning we are having strange behavior in hosts that have rebooted.

I have a host that went down for a scheduled reboot, and after it came back up none of its agent checks were reporting data. We simulated this with a second host by stopping traffic from the zabbix server to the host for 9 minutes then removed the block. The problem was duplicated on this host as well. I then tried a server with a different OS and was again able to duplicate the problem.

We have tried restarting the agents, upgrading the agents... but to no avail. All three servers are running agent 1.8.3 since we had just updated the host.

In trying to figure this out we did find 817 things in the Zabbix agent queue for over 10 minutes, some for months, and none of it ping.

Is anyone else having this problem? Thanks.

[Littlelink]

We do have the same Problem. After Upgrading the Server to 1.8.4, he do not any checks ( even if that there is an entry in the Log of Clients "zabbix_agentd active check started" ).

I can get all Elements via zabbix_get manually, but the Server do not update anything.

This is a really big problem and if i wont find a Solution fast we will have to go back to 1.8.3.

Additional we do get this one ( only for one Item of hundreds ), not often but it is in Server log ( Get value from agent failed: ZBX_TCP_READ() failed [Interrupted system call] ). On Client side there is nothing in the log.