4 Replies Latest reply on Nov 17, 2020 4:25 AM by Juan Gutierrez Ruiz

    Node Behavior on IDQ

    Raja Shahnawaz Soni Active Member

      Hello,

       

      We are facing an issue where both of the IDQ nodes are pointing to the same domain DB but then one of the nodes frequently goes down with the error:

      Cannot update the data for the master gateway. The domain configuration repository database may not be available. Verify that the database is available.

      com.informatica.jdbc.oraclebase.ddej: [informatica][Oracle JDBC Driver

       

      If it was the database issue then both the nodes should have gone down, why does only one of the node (the master one) go down frequently? Or how does the configuration behave typically, any official admin documentation for this behavior?

      Also how to deal with such scenarios in the future, any inputs are highly appreciated.

       

      Any insights here will be helpful.

        • 1. Re: Node Behavior on IDQ
          Nico Heinze Guru

          Honestly I would open a Service Request at Informatica Global Customer Support (GCS), this looks strange to me.

          Having written that, we also don't (yet) know anything about your setup. You did not mention which version of the Informatica platform / IDQ you work with, you haven't told us anything about the operating system and version of the two(?) nodes, and you did not tell us whether these nodes are located within the same physical subnet or not. If they belong to different subnets, then you should require your network admins to thoroughly check for any network drop-outs, because to me at first glance this looks like a network problem.

          So please give us a bit more information, maybe then we can help you better than with "stupid" suggestions like above.

           

          Regards,

          Nico

          • 2. Re: Node Behavior on IDQ
            Raja Shahnawaz Soni Active Member

            Dear Nico, thanks for your inputs, in fact, I should have shared more information here. So here it goes :

             

            1. RHEL 7 is the OS

            2. IDQ 10.2HF1

            3. Multinode setup with single domain DB for both the nodes

            4. Nodes are located on the same physical subnet (Although I will cross-check this once, but for now please consider it on the same physical subnet)

            • 3. Re: Node Behavior on IDQ
              Nico Heinze Guru

              Hi Raja,

               

              as long as we're talking about RHEL 7.3 or later (and still some 7.x version), that should work fine. At least as far as I recall, 7.3 onward are supported for 10.2 HF1.

               

              Both nodes are set up as gateway nodes, if I understand you correctly. So the fact that one node has this problem (and the other one does not show this behaviour) indicates to me either some network problem (maybe a problem with the network interface card) or some other technical issue.

               

              There might be a way to make the system work stably. There is a custom property named MasterDBRefreshInterval which you can set for the domain as a whole. The default value is 8000 milliseconds, in many cases it helps to increase this to e.g. 30000 milliseconds. You might want to give it a try.

               

              Otherwise I can only repeat myself here, please raise a service request if this doesn't help.

               

              Regards,

              Nico

              • 4. Re: Node Behavior on IDQ
                Juan Gutierrez Ruiz Active Member

                Hi Raja,

                 

                 

                In addition to Nico's responses, you should be aware of 2 additional parameters as you have a multi node environment. These parameters are infa.updateMasterConnectTimeout and infa.masterUpdateTimeInterval and they may need to be modified if MasterDBRefreshInterval is increased.

                 

                 

                Please perform following inequality when modifying these 3 values.

                (4 * MasterDBRefreshInterval + 45 + infa.updateMasterConnectTimeout) < (6 * infa.masterUpdateTimeInterval)

                 

                 

                Please check KB152445 for further details on how to set those parameters and select optimum values for them

                https://knowledge.informatica.com/s/article/152445

                 

                 

                If this does not solve the problem, please raise a case with Global Customer Support.

                 

                 

                Thanks,

                Juan

                1 of 1 people found this helpful