Skip to Content
0
Former Member
Mar 23, 2009 at 07:49 AM

Handling of the Enqueue Replication Server

499 Views

Hi all.

I installed a high availability system with a replicated enqueue server for the first time. The system is running but i have a question for understanding...

System constallation:

Database HA Cluster (HA Software HP serviceguard)

Application (also serviceguard): SCS and Web Dispatcher HA running on server one. The java central instance is also running on this machine. On the second cluster node a dialog instance is running.

For the Enqueue Replication Server we created the ERS-Instances on both servers (Server one:SCS, WD, JC; Server 2: DI).

When i start the system it appears in the following order:

Server A - SCS, ERS, JC

Server B - ERS, DI

My question is: Do i have to start the ERS on the SCS-Server? I think this step is necessary on the remote machine. Not on the SCS-Server... Or could there be a configuration error of myself?

At the moment i get this ERS-error message on the server where the SCS is running:

[Thr 47180764330720] Mon Mar 23 08:11:47 2009

[Thr 47180764330720] ***LOG Q0I=> NiIRead: recv (104: Connection reset by peer) [nixxi.cpp 4424]

[Thr 47180764330720] *** ERROR => NiIRead: SiRecv failed for hdl 0 / sock 8

+ (SI_ECONN_BROKEN/104; UD; ST; 10.1.237.125:55516) [nixxi.cpp 4424]+

[Thr 47180764330720] *** ERROR => RepServer: main: connect failed with return code -3

enrepserv.cp 748]

Thanks for your help.

Martin