Created attachment 924557 [details] server.log Description of problem: Error in server.log file after clean install and server start: Version-Release number of selected component (if applicable): JON 3.3 DR01 How reproducible: Steps to Reproduce: 1. Install and start JON 2. 3. Actual results: Warnings in server.log after clean install/start. Expected results: Should be zero silent warnings in server.log Additional info:
Closing this as the reporter did not take any time to clearly indicate the error. "Something is broken" is insufficient data"
Re-opening this bug as it is impacting our customers. Seems like we have overlooked this simple requirement. For txn to be isolated from each JBoss ON server in a cluster, each EAP server will have to have a unique identifier. This unique identifier needs to be defined for the transactions subsystem's core-environment node-identifier atttribute: ... <subsystem xmlns="urn:jboss:domain:transactions:1.4"> <core-environment node-identifier="${rhq.server.high-availability.name:rhq-server-01}"> ... If this is not defined, transaction recovery may be attempted on incorrect instances resulting in failures as identified in https://access.redhat.com/solutions/260023
The mentioned default of rhq-server-01 from comment 2 won't work as there are some noted restrictions to the node id value. For example, no hyphens and the maximum length is 28 characters. Meaning that we probably can't use the raw value provided by rhq.server.high-availability.name. Instead, we will need to remove hyphens and truncate to 28 characters and hope that it is still unique.
We have passed the deadline for 3.3 update 10 payload. A low risk fix for this issue will require more research so I am resetting the target to 3.3.11 for further consideration. The current assessment is that for new installations, setting the node-identifier would be possible. However, for existing installations, changes to the node-identifier could impact in-process transactions. This impact would not be limited to HA or customers who have experienced transactions issues. It would impact all installations, including single server. Granted, if the server finished its work on shutdown and there were no in-process transactions, there would be no issues. But the concern is with those environments that perhaps are performing ungraceful shutdowns or have large volumes of work left in the queue at shutdown.
In the master: commit d3375b4a69210e3b5dcc56aa4d9e697d67c67081 (HEAD -> master) Author: Michael Burman <miburman> Date: Wed Aug 15 15:12:58 2018 +0300 [BZ 1127359] Parameter jboss.tx.node.id in rhq-server.properties allows setting EAP HA identifier during installation
In the master: commit 8146a892f9c8174304cb43f64a7becef57d58243 (HEAD -> master) Author: Michael Burman <miburman> Date: Wed Sep 19 18:42:25 2018 +0300 [BZ 1139102] Update description for jboss.tx.node.id
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHSA-2018:2930