TrueSight Orchestration (TSO) adapter requests fail, regardless of the timeout specified in the request or configuration:

ERROR - [Current Time=Mon Jan 01 01:01:01 EDT 2014] [Process Name=:MyModule:Utilities:Perform Execute NSH Command] [Root Job Id=d4d7affcbb1eeeee:10b2bb83:13f9a79b182:-80001-1372701111454] [Job Id=d4d7affcbb1eeeee:10b2bb83:13f9a79b182:-80001-1372701111454/:MyModule:BladeLogic_Utilities:Get Compliance Job Run Results:start:call-process[2]/:MyModule:NSH:Execute NSH Command:start:call-process[1]]
An error occurred that triggered compensation:
Summary: Execution failed.

Caused by:
Summary: Message[summary=Timeout occurred., detail=A request was sent to a remote peer and a response was not received prior to the specified timeout.]

There are two solutions for this issue based on how long the adapter request runs before the error:

If your request runs for one hour, the issue is that the peers are timing out while waiting for each other. If a workflow running on one peer calls an adapter on another peer, there is a hard-coded one hour timeout before the peer running the workflow assumes the other peer is offline and stops waiting.

To extend this timeout, open or create the file [BAO_HOME]/config/tuning-config.xml and add the following XML:

<config>
    <adapter-manager>
        <adapter-call-timeout>3600000</adapter-call-timeout>
    </adapter-manager>
</config>

This parameter is in milliseconds; 3600000 milliseconds is one hour, the default. Change the value to a duration that works for your environment, save the file, and restart the peer. For more information about tuning-config.xml, see the article TrueSight Orchestration-Tuning-Parameters.

If your request runs for one minute, the peers are trying to look up the adapter configuration based on the name you provided but cannot because the adapter has been renamed. Restart the grid to clear its memory of the old name and associate adapter names with configurations correctly.