Created attachment 798836 [details] node one server log Attempt for redeploy of business central in cluster ended up with "Operation failed or was rolled back on all servers". Server logs contain exceptions with root cause: org.apache.helix.HelixException: instance nodeOne_12345 already has a liveinstance in cluster jbpm-cluster Version-Release number of selected component (if applicable): BPMS-6.0.0.ER3 Steps to Reproduce: 1. configure cluster with 2 nodes 2. deploy business central 3. undeploy business central 4. deploy business central again => operation failed, see server logs Attaching server logs from both nodes - containing useful information
Created attachment 798837 [details] node two server log
To make it clear - there is a workaround - to stop & destroy the cluster completely, create it again and deploy business central once more. But this can be hardly final solution.
main issues was that cluster member (helix) that is embedded in the application was not disconnecting itself on application shutdown, similar thing was for jbpm-executor which after application shut down was still trying to execute queries. fix applied in following repositories: kie-commons master: https://github.com/droolsjbpm/kie-commons/commit/f02ae07b84eb837050e5c684756ab34c47f50bc4 6.0.x: https://github.com/droolsjbpm/kie-commons/commit/0119d3bbd1b81b092089d574e828fc1d34d6eb96 jbpm: master:https://github.com/droolsjbpm/jbpm/commit/b775378260c547b802460fdb1d4e1b1759165808 6.0.x: https://github.com/droolsjbpm/jbpm/commit/18c8e5f95877e0f80d6670c73cacfdc41203205b kie-wb-distributions: master:https://github.com/droolsjbpm/kie-wb-distributions/commit/9d19c1474d2c44f595393d3439c6c1d33a397199 6.0.x: https://github.com/droolsjbpm/kie-wb-distributions/commit/bf5fff9f7ad511cabf817d6505d52f418e6e1297 jbpm-console-ng master:https://github.com/droolsjbpm/jbpm-console-ng/commit/5234a497c610d9de42075826ff5597ddef8a0eb2 6.0.x: https://github.com/droolsjbpm/jbpm-console-ng/commit/85f29dc6be17b7a9b30fea8ec4103644061c2e1b
Verified on BPMS-6.0.0.ER4