Ad Widget

Collapse

AIX - cannot read response: timed out

Collapse
X
 
  • Time
  • Show
Clear All
new posts
  • FreddyBeach
    Junior Member
    • Sep 2025
    • 5

    #1

    AIX - cannot read response: timed out

    I am starting to do a Proof of Concept with Zabbix, and have the server installed on RHEL 9.4 with a PostgreSQL DB.
    20+ windows servers being monitored, no issues.

    We have two AIX 7.2 LPAR monitored, with the 7.4.3-1 agent installed, with PSK no issues.

    Attempting to add a third AIX LPAR, that is an Oracle DB server, using same setup as the previous two AIX LPAR, but keep getting the Availability in RED for ZBX.
    Mousing over the icon we see, "Get value from agent failed: Cannot read response: timed out"

    I have increased the Timeout value from 3 to 10 , but issue still remains.

    Basic telnet to a port from the Zab Server to the LPAR to monitor is successful.

    Any ideas what to try next to debug this issue?

    Zabbix Agent: 7.4.3-1 on AIX 7.2
    Zabbix Server 7.4.5 on RHEL 9.4
  • hrrapozo
    Banned
    • Nov 2025
    • 25

    #2
    Is the Zabbix agent on the third AIX LPAR system enabled? Another thing to check is if that system's zabbix agent conf file is configured to accept connections from your zabbix server.

    Comment

    • FreddyBeach
      Junior Member
      • Sep 2025
      • 5

      #3
      Originally posted by hrrapozo
      Is the Zabbix agent on the third AIX LPAR system enabled? Another thing to check is if that system's zabbix agent conf file is configured to accept connections from your zabbix server.
      The Zabbix agent is running on the third system, and there is output in /var/log/zabbix/zabbix_agentd.log - Nothing obviously wrong shown in the log file.
      The Agent config file is identically to the other two AIX LPARs, only the PSK info was changed for the third LPAR. No issues with PSK , as it shows as green on the dashboard.

      Comment

      • FreddyBeach
        Junior Member
        • Sep 2025
        • 5

        #4
        Thus far I have:

        Set debug level to 4 both on the agent and the server, and could not find source of problem.
        Set debug on a good server to 4 and compared to the broken server at debug 4, couldn't find issue.

        Downloaded more recent version of agent, and re-did the agent install and config, still no date going to server.

        Compared resolv.conf and netsvc.conf , same contents.
        got our network guy to dig into tcpdump, and all looks ok

        What other AIX debugging can I try?
        What other permissions etc can I try?

        I followed most of this guide(recipe), but used the rpm for installation.



        Is there a checklist of the basic that I should confirm as next step(DNS, routing etc)?

        There LPAR was previously part of a HACMP cluster, and I wonder if there is still some remnants of that causing this issue.

        Comment

        • FreddyBeach
          Junior Member
          • Sep 2025
          • 5

          #5
          Figured it out.

          There was a /etc/resolv.conf file on this LPAR, and most of our servers don't have one. Then reviewing the /etc/netsvc.conf file , it was set to hosts=local, bind. I changed the file to ber hosts=local4, bind4 , and the zabbix agent can now communicate with the server.

          The question now is, how did this stop comms, as everything the zabbix agent.conf file is by IP ?

          Comment

          Working...