Bug 667983

Summary:	sendmail gets wedged
Product:	[Fedora] Fedora	Reporter:	Thomas J. Baker <tjb>
Component:	sendmail	Assignee:	Jaroslav Škarvada <jskarvad>
Status:	CLOSED NEXTRELEASE	QA Contact:	Fedora Extras Quality Assurance <extras-qa>
Severity:	high	Docs Contact:
Priority:	low
Version:	14	CC:	jskarvad, mlichvar
Target Milestone:	---
Target Release:	---
Hardware:	x86_64
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2012-01-31 13:49:20 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 Thomas J. Baker 2011-01-07 15:17:52 UTC

I'm running F14 with all updates and since going to 14, my sendmail-8.14.4-10.fc14.x86_64 dies at around 4:00am on a regular basis with the following error:


Jan  7 04:03:07 wintermute sendmail[800]: NOQUEUE: --- 451 4.0.0 opendaemonsocket: daemon MSA: cannot bind: Address already in use (hold)
Jan  7 04:03:07 wintermute sendmail[800]: NOQUEUE: SYSERR(root): opendaemonsocket: daemon MSA: cannot bind: Address already in use
Jan  7 04:03:07 wintermute sendmail[800]: daemon MSA: problem creating SMTP socket
Jan  7 04:03:12 wintermute sendmail[800]: NOQUEUE: --- 451 4.0.0 opendaemonsocket: daemon MSA: cannot bind: Address already in use (hold)
Jan  7 04:03:12 wintermute sendmail[800]: NOQUEUE: SYSERR(root): opendaemonsocket: daemon MSA: cannot bind: Address already in use
Jan  7 04:03:12 wintermute sendmail[800]: daemon MSA: problem creating SMTP socket
Jan  7 04:03:17 wintermute sendmail[800]: NOQUEUE: --- 451 4.0.0 opendaemonsocket: daemon MSA: cannot bind: Address already in use (hold)
Jan  7 04:03:17 wintermute sendmail[800]: NOQUEUE: SYSERR(root): opendaemonsocket: daemon MSA: cannot bind: Address already in use
Jan  7 04:03:17 wintermute sendmail[800]: daemon MSA: problem creating SMTP socket
Jan  7 04:03:22 wintermute sendmail[800]: NOQUEUE: --- 451 4.0.0 opendaemonsocket: daemon MSA: cannot bind: Address already in use (hold)
Jan  7 04:03:22 wintermute sendmail[800]: NOQUEUE: SYSERR(root): opendaemonsocket: daemon MSA: cannot bind: Address already in use
Jan  7 04:03:22 wintermute sendmail[800]: daemon MSA: problem creating SMTP socket
Jan  7 04:03:27 wintermute sendmail[800]: NOQUEUE: --- 451 4.0.0 opendaemonsocket: daemon MSA: cannot bind: Address already in use (hold)
Jan  7 04:03:27 wintermute sendmail[800]: NOQUEUE: SYSERR(root): opendaemonsocket: daemon MSA: cannot bind: Address already in use
Jan  7 04:03:27 wintermute sendmail[800]: daemon MSA: problem creating SMTP socket
Jan  7 04:03:32 wintermute sendmail[800]: NOQUEUE: --- 451 4.0.0 opendaemonsocket: daemon MSA: cannot bind: Address already in use (hold)
Jan  7 04:03:32 wintermute sendmail[800]: NOQUEUE: SYSERR(root): opendaemonsocket: daemon MSA: cannot bind: Address already in use
Jan  7 04:03:32 wintermute sendmail[800]: daemon MSA: problem creating SMTP socket
Jan  7 04:03:37 wintermute sendmail[800]: NOQUEUE: --- 451 4.0.0 opendaemonsocket: daemon MSA: cannot bind: Address already in use (hold)
Jan  7 04:03:37 wintermute sendmail[800]: NOQUEUE: SYSERR(root): opendaemonsocket: daemon MSA: cannot bind: Address already in use
Jan  7 04:03:37 wintermute sendmail[800]: daemon MSA: problem creating SMTP socket
Jan  7 04:03:42 wintermute sendmail[800]: NOQUEUE: --- 451 4.0.0 opendaemonsocket: daemon MSA: cannot bind: Address already in use (hold)
Jan  7 04:03:42 wintermute sendmail[800]: NOQUEUE: SYSERR(root): opendaemonsocket: daemon MSA: cannot bind: Address already in use
Jan  7 04:03:42 wintermute sendmail[800]: daemon MSA: problem creating SMTP socket
Jan  7 04:03:42 wintermute sendmail[800]: NOQUEUE: --- 421 4.0.0 opendaemonsocket: daemon MSA: server SMTP socket wedged: exiting (hold)
Jan  7 04:03:42 wintermute sendmail[800]: NOQUEUE: SYSERR(root): opendaemonsocket: daemon MSA: server SMTP socket wedged: exiting

Now I know this is when cron.daily runs, I do get over 100 logwatches sent to me at that time, and since it doesn't happen every night, it seems more load related than an easily reproducible bug. After it happened a few days ago, I turned log level up to 20 but it didn't produce any more enlightening error messages. 

Are there sendmail config file tweaks that may help?

Comment 1 Jaroslav Škarvada 2011-01-18 23:18:59 UTC

Do you limit number of sendmail child processes (confMAX_DAEMON_CHILDREN)? Are you using logwatch? Is there any specific sendmail settings in you config differing from the defaults?

I tried (several times) to handle more than 100 logwatch emails simultaneously on my testing machine - it was kicked into knees for couple of minutes but I was not successful to reproduce this.

Comment 2 Thomas J. Baker 2011-01-19 15:50:31 UTC

I have these set:

define(`confMAX_DAEMON_CHILDREN', `50')dnl
define(`confCONNECTION_RATE_THROTTLE', `3')dnl

I am using logwatch and I have a couple of dnsbl's defined as well and clam-av milter and spamass-milter configured. I receive a lot of mail on this machine.

Comment 3 Jaroslav Škarvada 2012-01-26 08:24:41 UTC

F14 is EOL, is it still an issue with current Fedoras?

Comment 4 Thomas J. Baker 2012-01-31 13:04:16 UTC

No, it's fine now.

Comment 5 Jaroslav Škarvada 2012-01-31 13:49:20 UTC

Thanks for info, closing according to comments 3, 4.