Description of problem: I got in server.log this line: 2014-03-19 15:53:31,273 ERROR [org.jboss.as.server.deployment.scanner] (DeploymentScanner-threads - 1) JBAS015052: Did not receive a response to the deployment operation within the allowed timeout period [60 seconds]. Check the server configuration file and the server logs to find more about the status of the deployment. Version-Release number of selected component (if applicable): current nightly master on fully updated fedora 19 engine+dwh+reports How reproducible: Not sure Steps to Reproduce: 1. install/setup engine+dwh+reports on a moderately-slow machine 2. 3. Actual results: Deployment fails Expected results: Should succeed Additional info: Increased the timeout to 10 minutes in [1], and it succeeds, took almost two minutes. [1] http://gerrit.ovirt.org/25895
Created attachment 876371 [details] server.log - failed after 60 seconds
Created attachment 876373 [details] server.log - succeeded after almost 2 minutes
I now checked logs of a 3.3 engine and deployment there took 4.5 minutes and was not stopped. Perhaps the default timeout changed, or something else in jboss.
Why is it downstream bug?
(In reply to Alon Bar-Lev from comment #5) > Why is it downstream bug? JBoss is not changed upstream.
Patch merged upstream on master branch for 3.5.0 rebase.
this was on fedora 19 which is not supported by rhevm, and also only on slow machine, which must be much slower than minimum hardware requirements for rhevm. we use rhevm on rhel on very slow machines every day and it works
(In reply to movciari from comment #10) > this was on fedora 19 which is not supported by rhevm, Don't think that's relevant, what's relevant is version of jboss - not sure what's happening there on the various OSes etc. > and also only on slow > machine, which must be much slower than minimum hardware requirements for > rhevm. IIRC the minimum hardware requirements do not specifically mention IO, and in any case, IO is much less predictable/controllable than cpu/memory, and imo we definitely want to make sure we don't fail, especially considering that it was an easy fix. > we use rhevm on rhel on very slow machines every day and it works Which version? With this patch applied, which is true for most versions of the last half year, you have ten minutes to deploy. If you want to reproduce, simply edit [1], change 'deployment-timeout="1200"' to something much lower, e.g. 'deployment-timeout="10"', and stop/start the engine. [1] /usr/share/ovirt-engine/services/ovirt-engine/ovirt-engine.xml.in
I do not think this requires doc text, as it did not affect, to the best of my knowledge, any released version.