Bug 1025765

Summary:	EBADF from select() cause TransportExceptions in client
Product:	[JBoss] JBoss Data Grid 6	Reporter:	Radim Vansa <rvansa>
Component:	CPP Client	Assignee:	Tristan Tarrant <ttarrant>
Status:	VERIFIED ---	QA Contact:	Alan Field <afield>
Severity:	medium	Docs Contact:
Priority:	unspecified
Version:	6.2.0	CC:	mgencur, sjacobs
Target Milestone:	CR2
Target Release:	6.2.1
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:		Doc Type:	Known Issue
Doc Text:	When more than several hundred Hot Rod client instances are created within a single process, TransportExceptions are thrown due to problems with connecting socket to the server. Mileage may vary depending on the OS configuration.	Story Points:	---
Clone Of:		Environment:
Last Closed:		Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:

Description Radim Vansa 2013-11-01 13:50:59 UTC

During stress testing, when about 200-300 clients per node are spawned, some of the sys::Socket::connect() attempts fail with TransportException.

This is preceded by the ::connect() call returning failing with EINPROGRESS, followed by ::select() failing with EBADF. However, even after the failure ::fnctl() reports the socket to be O_RDWR | O_NONBLOCK (as it should be) without error.

I had a suspicion that this may be related to the number of file descriptors, but ulimit -n shows that the limits are high above that.

Comment 2 Martin Gencur 2014-03-27 09:00:09 UTC

This was fixed a long time ago. Setting ON_QA

Comment 3 Alan Field 2014-03-31 14:45:40 UTC

Verified with JDG 6.2.1 CR3