[issue12210] test_smtplib: intermittent failures on FreeBSD

[Irit Katriel]

Irit Katriel  added the comment:

2.7-only issue.

--
nosy: +iritkatriel
resolution:  -> out of date
stage: needs patch -> resolved
status: open -> closed

___
Python tracker 

___
___
Python-bugs-list mailing list
Unsubscribe: 
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com

[issue12210] test_smtplib: intermittent failures on FreeBSD

[Stefan Krah]

Stefan Krah added the comment:

If it is a DNS failure, the timeout of 3s indeed looks too low for me.

On a misconfigured machine I'm easily getting timeouts of 20s, (not for
this particular test, just in general, e.g. when using ssh).

--

___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue12210
___
___
Python-bugs-list mailing list
Unsubscribe: 
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com

[issue12210] test_smtplib: intermittent failures on FreeBSD

[R. David Murray]

R. David Murray added the comment:

It would have to be a misconfigured machine that doesn't have 'localhost' in 
/etc/hosts (so that a lookup of localhost goes through DNS).  Or, I suppose, 
one that prioritizes the resolver over /etc/hosts (is that even possible?)

I would think other tests would fail if DNS has a 20 second latency.  I think 
having localhost configured in /etc/hosts is a requirement for other tests, so 
if that really is the problem I'm not sure we should be fixing the test suite 
here.  On the third hand, it doesn't really matter in this case and we're 
already using 15 for other tests, so I have no objection to making the test 
consistent.

--

___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue12210
___
___
Python-bugs-list mailing list
Unsubscribe: 
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com

[issue12210] test_smtplib: intermittent failures on FreeBSD

[Stefan Krah]

Stefan Krah added the comment:

I think it's also possible to misconfigure /etc/nsswitch.conf.

When I opened this issue, my buildbot had the FreeBSD default
configuration.  I no longer have any FreeBSD machines, so I
cannot check this now.

--

___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue12210
___
___
Python-bugs-list mailing list
Unsubscribe: 
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com

[issue12210] test_smtplib: intermittent failures on FreeBSD

[R. David Murray]

R. David Murray added the comment:

Those are in the second (or rather the first, in the file) group of tests, 
which have a timeout of 3 on python3.  The comment in the test case says that 
DNS lookup delays can cause timeouts.  Is there any chance that that machine 
sometimes has problems looking up 'localhost'?

There is no record in this issue of any change being applied to the tests, and 
I don't see any changes by Stefan in the file.

I suppose there's no real reason not to make all the timeouts 15 on both 2.7 
and python3.  It is weird that those tests should take so long to run, but more 
likely to be an environment problem than a python code problem.  On the other 
hand, you might want to look in to why this is happening on FreeBSD, it might 
reveal something interesting.  

The problem with this theory, of course, is that we haven't apparently seen the 
timeouts on python3.  But I'm not sure what to look at to try to diagnose a 
difference...the smtplib code isn't that different between the two, so 
presumably if there is a real difference it is at the level of the socket code. 
 Since this is as far as we know only a test problem, we should probably just 
go ahead and increase the timeouts.

--

___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue12210
___
___
Python-bugs-list mailing list
Unsubscribe: 
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com

[issue12210] test_smtplib: intermittent failures on FreeBSD

[koobs]

koobs added the comment:

Observed the following test_smtplib failure on koobs-freebsd10, noting that the 
timeout parameter is still 3 

test test_smtplib failed -- Traceback (most recent call last):
  File 
/usr/home/buildbot/python/2.7.koobs-freebsd10/build/Lib/test/test_smtplib.py, 
line 222, in testHELP
smtp = smtplib.SMTP(HOST, self.port, local_hostname='localhost', timeout=3)
  File /usr/home/buildbot/python/2.7.koobs-freebsd10/build/Lib/smtplib.py, 
line 256, in __init__
(code, msg) = self.connect(host, port)
  File /usr/home/buildbot/python/2.7.koobs-freebsd10/build/Lib/smtplib.py, 
line 317, in connect
(code, msg) = self.getreply()
  File /usr/home/buildbot/python/2.7.koobs-freebsd10/build/Lib/smtplib.py, 
line 365, in getreply
+ str(e))
SMTPServerDisconnected: Connection unexpectedly closed: timed out[283/401] 
test_abstract_numbers

Complete log attached.

--
nosy: +koobs
resolution: wont fix - 
stage: resolved - needs patch
status: closed - open
Added file: 
http://bugs.python.org/file39777/koobs-freebsd10.python.27.build978.log

___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue12210
___
___
Python-bugs-list mailing list
Unsubscribe: 
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com

[issue12210] test_smtplib: intermittent failures on FreeBSD

[Stefan Krah]

Stefan Krah added the comment:

Since this occurs only on 2.7 and *very* rarely, let's close it.

--
resolution: out of date - wont fix
status: open - closed

___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue12210
___
___
Python-bugs-list mailing list
Unsubscribe: 
http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com

[issue12210] test_smtplib: intermittent failures on FreeBSD

[Stefan Krah]

Stefan Krah stefan-use...@bytereef.org added the comment:

Naturally, as soon as I declare it fixed, it occurs again:

http://www.python.org/dev/buildbot/all/builders/AMD64%20FreeBSD%208.2%202.7/builds/326

--
status: closed - open

___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue12210
___
___
Python-bugs-list mailing list
Unsubscribe: 
http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com

[issue12210] test_smtplib: intermittent failures on FreeBSD

[Stefan Krah]

Stefan Krah stefan-use...@bytereef.org added the comment:

I haven't seen this in a while, so let's assume it's fixed.

--
resolution:  - out of date
stage:  - committed/rejected
status: open - closed

___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue12210
___
___
Python-bugs-list mailing list
Unsubscribe: 
http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com

[issue12210] test_smtplib: intermittent failures on FreeBSD

[Stefan Krah]

New submission from Stefan Krah stefan-use...@bytereef.org:

With Python 2.7, the FreeBSD AMD64 bot has sporadic failures in
test_smtplib. Since these failures don't occur in other branches,
I wonder if the timeout of 3 seconds is too low in 2.7. The timeout
is 15 seconds in 3.3.


test test_smtplib failed -- Traceback (most recent call last):
  File 
/usr/home/buildbot/buildarea/2.7.krah-freebsd/build/Lib/test/test_smtplib.py, 
line 186, in testNOOP
smtp = smtplib.SMTP(HOST, self.port, local_hostname='localhost', timeout=3)
  File /usr/home/buildbot/buildarea/2.7.krah-freebsd/build/Lib/smtplib.py, 
line 242, in __init__
(code, msg) = self.connect(host, port)
  File /usr/home/buildbot/buildarea/2.7.krah-freebsd/build/Lib/smtplib.py, 
line 303, in connect
(code, msg) = self.getreply()
  File /usr/home/buildbot/buildarea/2.7.krah-freebsd/build/Lib/smtplib.py, 
line 352, in getreply
raise SMTPServerDisconnected(Connection unexpectedly closed)
SMTPServerDisconnected: Connection unexpectedly closed



test test_smtplib failed -- Traceback (most recent call last):
  File 
/usr/home/buildbot/buildarea/2.7.krah-freebsd/build/Lib/test/test_smtplib.py, 
line 206, in testVRFY
smtp = smtplib.SMTP(HOST, self.port, local_hostname='localhost', timeout=3)
  File /usr/home/buildbot/buildarea/2.7.krah-freebsd/build/Lib/smtplib.py, 
line 242, in __init__
(code, msg) = self.connect(host, port)
  File /usr/home/buildbot/buildarea/2.7.krah-freebsd/build/Lib/smtplib.py, 
line 303, in connect
(code, msg) = self.getreply()
  File /usr/home/buildbot/buildarea/2.7.krah-freebsd/build/Lib/smtplib.py, 
line 352, in getreply
raise SMTPServerDisconnected(Connection unexpectedly closed)
SMTPServerDisconnected: Connection unexpectedly closed

--
components: Tests
keywords: buildbot
messages: 137217
nosy: r.david.murray, skrah
priority: normal
severity: normal
status: open
title: test_smtplib: intermittent failures on FreeBSD
type: behavior
versions: Python 2.7

___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue12210
___
___
Python-bugs-list mailing list
Unsubscribe: 
http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com

[issue12210] test_smtplib: intermittent failures on FreeBSD

[Stefan Krah]

Stefan Krah stefan-use...@bytereef.org added the comment:

Indeed, increasing the timeout to 15 solves the problem (I ran the
test with the -F option for about 100 times).

Since the timeout has been increased before to deal with flaky tests
(52682bd738a7), I guess it's ok to do the same for 2.7.

--
keywords: +patch
Added file: http://bugs.python.org/file22181/issue12210.diff

___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue12210
___
___
Python-bugs-list mailing list
Unsubscribe: 
http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com

[issue12210] test_smtplib: intermittent failures on FreeBSD

[Stefan Krah]

Stefan Krah stefan-use...@bytereef.org added the comment:

Hmm, testVRFY occurs twice, once with timeout=15 and once with
timeout=3. So there might be another reason why the test passes on
3.3 but not on 2.7.

Anyway, increasing the timeout to 15 throughout does solve
(or suppress) the problem.

--

___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue12210
___
___
Python-bugs-list mailing list
Unsubscribe: 
http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com