Bug 812452

Summary:	[eap6] timeout issues - deployment
Product:	[Other] RHQ Project	Reporter:	Libor Zoubek <lzoubek>
Component:	Agent	Assignee:	Heiko W. Rupp <hrupp>
Status:	CLOSED CURRENTRELEASE	QA Contact:	Mike Foley <mfoley>
Severity:	unspecified	Docs Contact:
Priority:	high
Version:	4.4	CC:	hrupp, theute
Target Milestone:	---
Target Release:	RHQ 4.4.0
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2013-09-01 10:16:56 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:
Bug Depends On:
Bug Blocks:	707223, 735475

Description Libor Zoubek 2012-04-13 18:51:07 UTC

Description of problem: In general, almost all operations invoked on EAP by agent can timeout. Today I had issues with creating deployment. My machine had swap filled by 700M  and EAP became too slow then.

This BZ might be related to other operations that are adding/removing a child on EAP.

Version-Release number of selected component (if applicable):


How reproducible:hard


Steps to Reproduce:
1.Have an overloaded machine - low memory
2.standup EAP6 on it, inventory it
3.create WAR Deployment on it using JON UI
  
Actual results:

If EAP is too slow, child resource creation fails with timeout. Although, deployment appears on EAP. Further attempts to create a resource will be refused by EAP (duplicate resource). But deployment is not discovered (operation failed and discovery was not triggered) so user expects it really failed.

Expected results: if deploying times out and fails, deployment should not be on EAP. Otherwise deployment must succeed.


Additional info: This can be much bigger issue not only for overloaded machine, but also larger deployments. I was deploying WAR that contains 1 JSP. Customers will more likely deploy larger applications and could potentially run into same issue.

Comment 1 Charles Crouch 2012-04-16 15:25:28 UTC

Potentially related to https://bugzilla.redhat.com/show_bug.cgi?id=802796

Comment 2 Heiko W. Rupp 2012-04-25 04:59:46 UTC

There are two issues to this:

- When you create a deployment, and do not specify a timeout explicitly then a value of 60 seconds is assumed, after which the plugin container reports a failure no matter what happens inside the plugin

- the plugin currently had a fixed timeout of 10s for connecting and 60s for upload.
-- the plugin does not have access to the timeout from the wizard to e.g. set its value to (x-1)s

One could pass the timeout from the UI to the plugin and give the plugin a hint about the minimal timeout requested from the user. Just setting it to e.g. 10min inside the plugin will not work, if the user does not specify a timeout in the UI.

Comment 3 Heiko W. Rupp 2012-04-25 09:01:12 UTC

master 38935c2

Default timeout for uploads is now increased to 120 seconds. The user can override this via timeout setting in the UI; this timeout setting applies to the total time the plugin is working in the createChild method. This means that with a setting of 120sec in the UI, it is possible that the actual time that an upload may take is only 119sec.

Comment 4 Heiko W. Rupp 2013-09-01 10:16:56 UTC

Bulk closing of items that are on_qa and in old RHQ releases, which are out for a long time and where the issue has not been re-opened since.