I have some questions regarding BizTalk 2004 and failover for running orchestrations that I hope someone can answer:
We have two BizTalk 2004 servers (running on Windows 2003 Server) in a BizTalk server group deployed with the same orchestrations. The SQL server is clustered on a Windows 2003 Cluster running on two different servers.
As we understand BizTalk 2004 is in an active-active configuration by design, without clustering. My question is how BizTalk handles a failover on one BizTalk server when an orchestration instance is started.
In our tests, the running orchestration instance on the failover server (the failover was done by rebooting the BizTalk server) was not continued on the other BizTalk server while the server failed. When the BizTalk server that failed was idle again the orchestration instance was continued on the same server as before.
Is the other BizTalk server supposed to continue the orchestration instance when the BizTalk server that was running the orchestration instance failed?
What happens if the failed BizTalk server never gets idle again, is the message lost?
Are there any settings to get this functionality?
Or maybe we have misunderstood how the active-active configuration works?
We have two BizTalk 2004 servers (running on Windows 2003 Server) in a BizTalk server group deployed with the same orchestrations. The SQL server is clustered on a Windows 2003 Cluster running on two different servers.
As we understand BizTalk 2004 is in an active-active configuration by design, without clustering. My question is how BizTalk handles a failover on one BizTalk server when an orchestration instance is started.
In our tests, the running orchestration instance on the failover server (the failover was done by rebooting the BizTalk server) was not continued on the other BizTalk server while the server failed. When the BizTalk server that failed was idle again the orchestration instance was continued on the same server as before.
Is the other BizTalk server supposed to continue the orchestration instance when the BizTalk server that was running the orchestration instance failed?
What happens if the failed BizTalk server never gets idle again, is the message lost?
Are there any settings to get this functionality?
Or maybe we have misunderstood how the active-active configuration works?