Inbound Singleton Adapters are Not Failing Over to a Standby Node After the Primary Node Crashes
Last updated on NOVEMBER 03, 2016
Applies to:Oracle SOA Suite - Version 184.108.40.206.0 and later
Oracle(R) BPEL Process Manager 10g - Version 10.1.3.4 and later
Information in this document applies to any platform.
***Checked for relevance on 09-Mar-2012***
You have an inbound singleton adapter(s)** deployed to a cluster, configured so that polling only takes place in one node of the cluster at a time.
You've found that when the active node is stopped abruptly, the surviving node of the cluster does not take over polling as it should.
In order to get the surviving node to poll, you have to go into the BPELConsole and toggle each process from Off to On.
If the active node is stopped gracefully (ie.. through the AS Console or the use of opmnctl commands), the surviving node does take over polling as expected. This issue only occurs when the active node is stopped abruptly.
Assuming a two node cluster, when both nodes are running, the polling for the adapter takes place in only one node. When the active node (which is doing the polling) is suddenly stopped, it's expected that the surviving node will take over and start to poll for the adapters.
An example of an abrupt stop of the active node on Linux\Unix is an execution of the operating system command:
**Commonly used singleton adapters are File and FTP
Sign In with your My Oracle Support account
Don't have a My Oracle Support account? Click to get started
My Oracle Support provides customers with access to over a
Million Knowledge Articles and hundreds of Community platforms