Created attachment 1222361 [details] engine log Description of problem: After running engine-setup after upgrade from 4.0.5-7 to 4.0.6-1 the setup fails on checking if dwhd is running even though dwhd service is not running and should have been stopped. DWH is local, not remote. Version-Release number of selected component (if applicable): 4.0.6-1 How reproducible: always Steps to Reproduce: 1. install 4.0.5-7 2. upgrade packages to 4.0.6-1 3. run engine-setup Actual results: fail on dwhd is running check Expected results: successful engine-setup Additional info: [ INFO ] Stopping dwh service [ INFO ] Stopping Image I/O Proxy service [ INFO ] Stopping websocket-proxy service [ ERROR ] dwhd is currently running. Its hostname is pm-rh40.rhev.lab.eng.brq.redhat.com. Please stop it before running Setup. [ ERROR ] Failed to execute stage 'Transaction setup': dwhd is currently running
Please attach postgresql logs. Most likely this is a duplicate of bug 1286441.
However, considering the number of duplicates it has, perhaps we should try to do something.
And please attach also dwhd logs. Thanks.
Created attachment 1222379 [details] dwhd log
Created attachment 1222380 [details] postgresql logs
Please also yum logs. Sorry :-( Thanks.
Created attachment 1222397 [details] yum log
Please attach also setup logs. Thanks.
Created attachment 1222399 [details] engine setup log ...I should have packed whole /var/log folder...
1. yum updated postgresql. yum.log: Nov 21 10:15:35 Updated: postgresql-server-9.2.18-1.el7.x86_64 2. dwhd fails to reconnect. First error in dwhd log: 2016-11-21 10:18:47|sIAAti|yb0VIK|q0Ha1J|OVIRT_ENGINE_DWH|OsEnumUpdate|Default|6|Java Exception|tJDBCInput_4|org.postgresql.util.PSQLException:FATAL: terminating connection due to administrator command|1 Seems like it never succeeded. 3. engine-setup stops dwhd "successfully": 2016-11-21 10:35:13 DEBUG otopi.plugins.otopi.services.systemd plugin.executeRaw:863 execute-result: ('/bin/systemctl', 'stop', 'ovirt-engine-dwhd.service'), rc=0 Nothing in stdout/stderr. 4. engine-setup checks the db and sees it's still "up": 2016-11-21 10:35:13 DEBUG otopi.ovirt_engine_setup.engine_common.database database.execute:222 Result: [{'var_value': '1', 'var_datetime': None, 'var_name': 'DwhCurrentlyRunning'}] Shirly, please have a look. Indeed seems like a duplicate of bug 1286441. But the fix there was supposed to work in this case - between upgrading PG and stopping dwhd passed 20 minutes, this should have been enough for dwhd to reconnect. It might be related to the upgrade. IIRC all the duplicates there are upgrades. Perhaps dwhd with the old pg client library can't connect to the upgraded pg server.
This issue must have been related to some particular setup and I can't reporoduce this anymore.