Report information
The Basics
Id:
46516
Status:
resolved
Priority:
Low/Low
Queue:
bind9-public
People
BugTracker
Version Fixed:
9.9.12, 9.10.7, 9.11.3, 9.12.0
Version Found:
(no value)
Versions Affected:
(no value)
Versions Planned:
(no value)
Priority:
P3 Low
Severity:
S3 Low
CVSS Score:
(no value)
CVE ID:
(no value)
Component:
(no value)
Area:
test
Dates
Created:Tue, 07 Nov 2017 03:05:34 -0500
Updated:Wed, 08 Nov 2017 03:34:24 -0500
Closed:Wed, 08 Nov 2017 03:34:24 -0500

This bug tracker is no longer active.

Please go to our Gitlab to submit issues (both feature requests and bug reports) for active projects maintained by Internet Systems Consortium (ISC).

Due to security and confidentiality requirements, full access is limited to the primary maintainers.

History
  Tue, 07 Nov 2017 03:05:35 -0500 Michał Kępień <michal@isc.org> - Ticket created
Date: Tue, 07 Nov 2017 09:05:34 +0100
From: michal@isc.org
To: bind9-public@isc.org
Subject: Stabilize notify system test

The notify system test sometimes fails on modern Linux systems [1] due to an already known kernel bug getting triggered by more than one thread attempting to send data using the same UDP socket (see RT #32355). This results in one of the threads seeing -EPERM returned for its sendmsg() call. Depending on which thread that turns out to be, the system test may fail due to one of the NOTIFY messages not being sent out in the last check ("checking notify to multiple views using tsig"). [1] https://jenkins.isc.org/view/BIND/job/bind9-master-fedora25-amd64-1/334/ https://jenkins.isc.org/view/BIND/job/bind9-master-fedora25-amd64-1/336/ https://jenkins.isc.org/view/BIND/job/bind9-master-centos7-amd64-1/338/ https://jenkins.isc.org/view/BIND/job/bind9-master-fedora25-amd64-1/341/ https://jenkins.isc.org/view/BIND/job/bind9-master-fedora26-64/136/
  Tue, 07 Nov 2017 03:09:39 -0500 Michał Kępień <michal@isc.org> - Correspondence added

The "race" to be the first thread using a socket is caused by the slave views inheriting "notify yes;" from global configuration. It causes a NOTIFY to be immediately sent back to view "a" from views "b" and "c" after view "a" notifies views "b" and "c" about a zone update. As these duplicate NOTIFY messages serve no purpose in this check, disabling them will stabilize the test by working around the kernel bug mentioned. While investigating this issue, I also noticed that junk gets logged when using "also-notify { <ip> key <keyname>; };". This is caused by a misplaced call to isc_sockaddr_format(). Please review rt46516. It fixes both problems mentioned above. IMHO, these fixes should be merged as two separate commits without a CHANGES note, but if anyone feels otherwise, I have no strong feelings about it, so please just let me know.
Tue, 07 Nov 2017 03:09:40 -0500 Michał Kępień <michal@isc.org> - Status changed from 'open' to 'review'
Tue, 07 Nov 2017 03:09:40 -0500 Michał Kępień <michal@isc.org> - Untaken
Tue, 07 Nov 2017 03:09:41 -0500 Michał Kępień <michal@isc.org> - Reference to #32355: added
Tue, 07 Nov 2017 23:43:18 -0500 Mark Andrews <marka@isc.org> - Taken
  Tue, 07 Nov 2017 23:49:01 -0500 Mark Andrews <marka@isc.org> - Correspondence added

looks fine.
Tue, 07 Nov 2017 23:49:02 -0500 Mark Andrews <marka@isc.org> - Given to Michał Kępień <michal@isc.org>
  Wed, 08 Nov 2017 03:34:23 -0500 Michał Kępień <michal@isc.org> - Correspondence added

Applied without CHANGES note as separate commits cherry-picked into master and all maintenance branches. 9.9.12, 9.10.7, 9.11.3, 9.12.0
Wed, 08 Nov 2017 03:34:24 -0500 Michał Kępień <michal@isc.org> - Status changed from 'review' to 'resolved'
Wed, 08 Nov 2017 03:34:24 -0500 Michał Kępień <michal@isc.org> - Version Fixed 9.9.12, 9.10.7, 9.11.3, 9.12.0 added