On 18/01/2007 at 09:06 +0100, Adrian von Bidder wrote:

> This is about the postgrey hang that I talked about previously.  Really
> Annoying. (And I'm still happily running postgrey without any 
> problems on my own machine, of course...)

Hi, I do not know if this is a regression or what. But since Tuesday I’m
experiencing this postgrey crashes (the same error than José Carlos).

This is the first time it happened on my server:

| Sep 11 02:55:31 bacterio postfix/smtpd[32191]: connect from 
nixe.bergantells.net[213.239.199.107]
| Sep 11 02:55:31 bacterio postgrey[19113]: action=pass, reason=client 
whitelist, client_name=nixe.bergantells.net, client_address=213.239.199.107, 
[...]
| Sep 11 02:55:31 bacterio postgrey[19113]: cleaning up old logs...
| Sep 11 02:55:31 bacterio postgrey[19113]: cleaning up old entries...
| Sep 11 02:55:31 bacterio postgrey[19113]: cleaning main database finished. 
before: 597, after: 500
| Sep 11 02:55:31 bacterio postgrey[19113]: FATAL: Can't call method 
"txn_commit" on an undefined value at /usr/sbin/postgrey line 289.
| Sep 11 02:55:32 bacterio postfix/smtpd[32191]: warning: connect to 
127.0.0.1:60000: Connection refused

And tonight again:

| Sep 12 02:39:42 bacterio postfix/smtpd[21792]: connect from 
89.1.126.105.dynamic.barak-online.net[89.1.126.105]
| Sep 12 02:39:43 bacterio postgrey[7560]: action=greylist, reason=new, 
client_name=89.1.126.105.dynamic.barak-online.net, client_address=89.1.126.105, 
[...]
| Sep 12 02:39:43 bacterio postgrey[7560]: cleaning up old logs...
| Sep 12 02:39:43 bacterio postgrey[7560]: cleaning up old entries...
| Sep 12 02:39:43 bacterio postgrey[7560]: cleaning main database finished. 
before: 138, after: 138
| Sep 12 02:39:43 bacterio postgrey[7560]: FATAL: Can't call method 
"txn_commit" on an undefined value at /usr/sbin/postgrey line 289.
| Sep 12 02:39:44 bacterio postfix/smtpd[21792]: warning: connect to 
127.0.0.1:60000: Connection refused

After it happened the first time yesterday, I wiped the postgrey
database (so hi doubt it would be db corruption).

> Jose, can you please confirm the versions of the postgrey, 
> libberkeleydb-perl, libdb4.3, libdb4.4 (depending on the libberkeleydb-perl
> verison either of those is used)  and perl Debian packages that are 
> installed on your system?

These are the versions of the packages you asked Jose:

| # dpkg-query -W postgrey libberkeleydb-perl libdb4.3 libdb4.4 perl
| libberkeleydb-perl      0.31-1
| libdb4.3        4.3.29-9
| libdb4.4        4.4.20-9
| perl    5.8.8-7
| postgrey        1.31-1

| # uname -a
| Linux bacterio 2.6.22.6 #1 Fri Aug 31 13:16:16 CEST 2007 i686 GNU/Linux

> Anything special about your system?  Xen, vmware or whatever?  NFS or
> some other network based filesystem on the partition where the database
> is located?

Kernel is a vanilla (from kernel.org) compiled by myself, I do not use
Xen and the db resides on a normal ide disk on the local machine.

I kwnow it’s not too much information I’m providing, as this is highly
annoying I would be very glad to test anything or provide any additional
information you should need.

Thanks in advance.

-- 
Kiko


Reply via email to