Re: [fetchmail-users] strange polling errors

[Matthias A. <mat...@gm...>]

Am 25.02.2011 15:52, schrieb Jonathan Buschmann:
> I've had the following errors in my syslog for some time now - whatever
> the problem it doesn't _seem_ to affect mail fetching though.
> I'm fetching from Exchange (wish I could find out its version).
> I'm using 6.3.19+NTLM+SSL on a SUN with Solaris 10.
> My fetchmailrc:
> set daemon 10
> set syslog
> set postmaster "postmaster"
> set no bouncemail
> set no spambounce
> set properties ""
> poll mail.<snip> with proto IMAP with auth password
>         user '<snip>' there with password '<snip>' is '<snip>' here
> options fetchall no mimedecode ssl idle
> 
> As you can see there is a very regular pattern, happening every 5:14,
> more or less. (the "awakened" is just a cron job that makes sure
> fetchmail is running)
> Feb 25 14:24:43 jnoblade fetchmail[1588]: [ID 676216 mail.error] re-poll
> failed
> Feb 25 14:24:43 jnoblade fetchmail[1588]: [ID 831971 mail.error] socket
> error while fetching from <snip>
> Feb 25 14:24:43 jnoblade fetchmail[1588]: [ID 991256 mail.info] Query
> status=2 (SOCKET)

Hi Jonathan,

Probably some docket timeout of ~300s.

Can I see a verbose trace?  Please run:

fetchmail -q
env LC_ALL=C fetchmail -vv

and check syslog after two "re-poll failed".

This should help us closing in on the whats and why it's failing.

Thanks.

Best regards
Matthias