Bug 1389093

Summary: [GSS](6.4.z) Session contention when node fails
Product: [JBoss] JBoss Enterprise Application Platform 6 Reporter: Shay Matasaro <smatasar>
Component: mod_clusterAssignee: Paul Ferraro <paul.ferraro>
Status: CLOSED WONTFIX QA Contact: Michal Karm Babacek <mbabacek>
Severity: high Docs Contact:
Priority: unspecified    
Version: 6.4.4CC: bmaxwell, bperkins, jawilson, jclere, paul.ferraro, pmackay, pslavice, rbost, rnetuka
Target Milestone: CR1   
Target Release: One-off release   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of:
: 1394057 (view as bug list) Environment:
Last Closed: 2019-08-08 16:08:04 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:
Bug Depends On:    
Bug Blocks: 1394057    

Description Shay Matasaro 2016-10-26 20:39:41 UTC 
This issue is related to both EAP and mod_cluster.

Scenario:
- An LB that sits in front of a few EAP nodes(EAP1, EAP2,EAP3, EAP4)
- A browser sending multiple requests(possibly concurrent) to EAP1
- EAP1 has an issue and stops responding
- LB gets request R1 from browser , and does not recognize the jvmroute(since EAP1 is down)
- LB redirects the request to EAP2
- The browser sends another request(R2) still with a jvm route that contains EAP1
- LB looks at R2 sees a faulty route and redirects it to EAP3
- Similar thing happens for R3 which gets sent to EAP4
- EAP2,EAP3,EAP4 all try to lock and own the session and most likely failing causing continuous retries and timeouts
Comment 1 Shay Matasaro 2016-11-02 11:49:53 UTC 
https://bugzilla.redhat.com/show_bug.cgi?id=1366939
https://bugzilla.redhat.com/show_bug.cgi?id=1366754
Comment 3 Paul Ferraro 2016-11-10 16:02:35 UTC 
https://github.com/modcluster/mod_cluster/commit/01cdb9b48c5788c92bf631edf1c288f49ef44c53