Bug 830336

Summary:	db deadlock return should not log error
Product:	Red Hat Enterprise Linux 6	Reporter:	Nathan Kinder <nkinder>
Component:	389-ds-base	Assignee:	Rich Megginson <rmeggins>
Status:	CLOSED ERRATA	QA Contact:	Sankar Ramalingam <sramling>
Severity:	unspecified	Docs Contact:
Priority:	medium
Version:	6.4	CC:	jgalipea, jrusnack
Target Milestone:	rc
Target Release:	---
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:	389-ds-base-1.2.11.12-1.el6	Doc Type:	Bug Fix
Doc Text:	Cause: The directory server is under a heavy read and write load, and an update request is processed. Consequence: An error message appears in the errors log: "entryrdn-index - _entryrdn_put_data: Adding the parent link (XXX) failed: DB_LOCK_DEADLOCK: Locker killed to resolve a deadlock (-30994)". Or other similar DB_LOCK_DEADLOCK error messages. Fix: Deadlocks are quite normal under concurrent write loads. The fix is to ensure deadlocks are handled properly (by retrying the transaction) and not to log these messages since they are not errors. Result: Directory Server can sustain a heavy write load without logging deadlock messages to the errors log.	Story Points:	---
Clone Of:		Environment:
Last Closed:	2013-02-21 08:17:27 UTC	Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description Nathan Kinder 2012-06-08 21:15:12 UTC

This bug is created as a clone of upstream ticket:
https://fedorahosted.org/389/ticket/345

When entryrdn gets a deadlock, it should not report an error at the default log level e.g.

[15/Apr/2012:07:33:33 -0400] entryrdn-index - _entryrdn_put_data: Adding the
parent link (P28354) failed: DB_LOCK_DEADLOCK: Locker killed to resolve a
deadlock (-30994)

It should log this at the TRACE log level.

Comment 1 Jenny Severance 2012-06-19 16:43:42 UTC

Please add steps to reproduce this dead lock.

Comment 2 Rich Megginson 2012-07-07 14:30:19 UTC

Covered by the new test multi_plugin - the stress test
testcases/DS/6.0/multi_plugin
------------------------------------------------------------------------
r6706 | rmeggins | 2012-07-07 08:24:23 -0600 (Sat, 07 Jul 2012) | 16 lines

Add multi_plugin test suite
This test enables and configures multiple plugins in the style
of ipa.  The stress version of this test will enable the 
TXN_TESTING in the server which will exercise the txn deadlock
and retry code paths in the server.
Tickets covered by this test:
Ticket #345 - db deadlock return should not log error
Ticket #387 - managed entry sometimes doesn't delete the managed entry
Ticket #360 - ldapmodify returns Operations error
Ticket #335 - transaction retries need to be cache aware
Bugs covered by this test:
830336 db deadlock return should not log error
830343 managed entry sometimes doesn't delete the managed entry
833218 ldapmodify returns Operations error
833202 transaction retries need to be cache aware


Steps to reproduce
1) run the multi_plugin stress tests
2) make sure no DB_LOCK_DEADLOCK messages are in the errors log

Comment 4 Ján Rusnačko 2013-01-29 11:31:58 UTC

On RHEL 6.4 Snapshot 5 with 389-ds-base-1.2.11.15-11.el6.x86_64:

multi_plugin startup 	100% (1/1)
multi_plugin run 	100% (6/6)
multi_plugin cleanup 	100% (1/1)

No errors in log.

Verified.

Comment 6 errata-xmlrpc 2013-02-21 08:17:27 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-2013-0503.html