Bug 903279

Summary:	plumage collector plugin crashes view server if it can't contact negotiator
Product:	Red Hat Enterprise MRG	Reporter:	Pete MacKinnon <pmackinn>
Component:	condor-plumage	Assignee:	Pete MacKinnon <pmackinn>
Status:	CLOSED ERRATA	QA Contact:	Stanislav Graf <sgraf>
Severity:	medium	Docs Contact:
Priority:	medium
Version:	2.3	CC:	dahorak, eerlands, matt, mkudlej, pmackinn, sgraf
Target Milestone:	2.5
Target Release:	---
Hardware:	All
OS:	Linux
Whiteboard:
Fixed In Version:	condor-7.8.9-0.7	Doc Type:	Bug Fix
Doc Text:	Cause: The plumage view-server plugin did not correctly detect a null socket, caused by failure to contact the negotiator. Consequence: Attempt to de-reference null socket caused a crash in the plugin. Fix: Plugin logic was fixed so that null sockets are detected and handled gracefully. Result: Plumage plugins no longer crash when the negotiator cannot be contacted.	Story Points:	---
Clone Of:
Clones:	1089287 (view as bug list)		Environment:
Last Closed:	2014-04-28 16:46:41 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:	1089287

Description Pete MacKinnon 2013-01-23 16:03:28 UTC

If the Plumage view server plugin cannot connect to the negotiator for obtaining userprio data, it crashes with a stack like the following:

01/23/13 15:58:06 Accumulating data: Time=1358953086
01/23/13 15:58:07 Can't find address for negotiator 
01/23/13 15:58:07 ODSAccountant: Can't connect negotiator for Accountant ad!
01/23/13 15:58:07 ODSAccountant: failed to send GET_PRIORITY command to negotiator!
Stack dump for process 18775 at timestamp 1358953087 (10 frames)
/usr/lib64/libcondor_utils_7_8_7.so(dprintf_dump_stack+0x131)[0x389633eca1]
/usr/lib64/libcondor_utils_7_8_7.so[0x3896385a22]
/lib64/libpthread.so.0[0x3632c0f500]
/usr/lib64/condor/plugins/PlumageCollectorPlugin-plugin.so(_ZN7plumage3etl13ODSAccountant7fetchAdEv+0x64)[0x7f0d78074934]
/usr/lib64/condor/plugins/PlumageCollectorPlugin-plugin.so(_ZN22PlumageCollectorPlugin18recordAccountantAdEv+0x50)[0x7f0d780747e0]
/usr/lib64/libcondor_utils_7_8_7.so(_ZN12TimerManager7TimeoutEPiPd+0x1a1)[0x38964478c1]
/usr/lib64/libcondor_utils_7_8_7.so(_ZN10DaemonCore6DriverEv+0x763)[0x3896455083]
/usr/lib64/libcondor_utils_7_8_7.so(_Z7dc_mainiPPc+0xf50)[0x38964445d0]
/lib64/libc.so.6(__libc_start_main+0xfd)[0x363241ecdd]
condor_collector[0x40d9f9]

The code needs to return from the fetchAd method instead of trying to go ahead and decode from the NULL socket.

Comment 3 Martin Kudlej 2013-06-20 09:17:06 UTC

I've found this bug in our long term condor instance. If problem occurs it is not possible to contact collector by condor_status.

Comment 9 Stanislav Graf 2014-03-10 14:15:25 UTC

I have checked that the patch is part of condor-7.8.9-0.7.el6 source rpm.

Our test suite using condor-plumage and mongodb works as before with this change. No issues found.

--> VERIFIED

Comment 11 errata-xmlrpc 2014-04-28 16:46:41 UTC

Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

http://rhn.redhat.com/errata/RHSA-2014-0440.html