You are viewing a plain text version of this content. The canonical link for it is here.
Posted to server-user@james.apache.org by Nathan Cheng <na...@yahoo.com> on 2005/05/14 09:30:46 UTC
query.bondedsender.org & 1000s of TIME_WAITs
Has anybody else noticed that query.bondedsender.org is apparently no longer functioning? At least it is not responding
to ping, and apparently is the cause for James' SMTP service not working for us the entire day: SMTP is working for the
first time today (technically yesterday--it's been a long "James" day) now that I've commented out the following from
our James config file:
<!-- People on this list agree to pay a penalty if they send spam
<mailet match="InSpammerBlacklist=query.bondedsender.org"
class="ToProcessor">
<processor> transport </processor>
</mailet>
-->
I'm still not sure that this is the problem though, since netstat is still returning thousands of these lines:
tcp 0 0 10.235.221.9:46882 10.235.221.12:mysql TIME_WAIT
and it just seems like it's not right to have 3118 of these at the same time, dying and being created seemingly by the
hundreds per second.
But SMTP is working now that bondedsender is no longer being consulted. But maybe that's just because the load has died
down and we still have an unsolved problem some where.
Any ideas anyone?
Thanks,
Nathan
---------------------------------------------------------------------
To unsubscribe, e-mail: server-user-unsubscribe@james.apache.org
For additional commands, e-mail: server-user-help@james.apache.org
Re: query.bondedsender.org & 1000s of TIME_WAITs
Posted by Nathan Cheng <na...@yahoo.com>.
I checked our config, and we have:
<data-source name="maildb" class="org.apache.james.util.dbcp.JdbcDataSource">
<driver>org.gjt.mm.mysql.Driver</driver>
<dburl>jdbc:mysql://somedomain/mail?autoReconnect=true</dburl>
<user>someuser</user>
<password>somepassword</password>
<max>50</max>
</data-source>
So I think the problem lies somewhere other than in the max_idle setting.
Nathan
apache@bago.org wrote:
>>>netstat is still returning thousands of these lines:
>>> tcp 0 0 10.235.221.9:46882 10.235.221.12:mysql TIME_WAIT
>>
>>Something is definitely not right, especially since JAMES
>>will use a connection pool, and a very limited # of JDBC connections.
>
>
> I think there is a problem with the default DBCP configuration of the
> max_idle property.
> I see that James by default set "max_idle = 0": I added
> <max_idle>80</max_idle> to my datasource configuration and seems to work
> better. (I have <max>80</max> in the same datasource configuration).
>
> I think the "max_idle = 0" should be considered a bug (IIRC dbcp by default
> use max_idle = 8, but james overwrite this with its default): can someone
> confirm my diagnosis?
>
> Stefano
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: server-user-unsubscribe@james.apache.org
> For additional commands, e-mail: server-user-help@james.apache.org
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: server-user-unsubscribe@james.apache.org
For additional commands, e-mail: server-user-help@james.apache.org
RE: query.bondedsender.org & 1000s of TIME_WAITs
Posted by "Noel J. Bergman" <no...@devtech.com>.
Stefano Bagnara wrote:
> I think there is a problem with the default DBCP configuration
> of the max_idle property.
That's possible. I did run:
svn log -r133208:155410 \
http://svn.apache.org/repos/asf/jakarta/commons/proper/dbcp/trunk
and the differences between DBCP 1.2.1 (r132208) and trunk (r155410) do not
appear significant to this issue.
> I see that James by default set "max_idle = 0"
That does not seem appropriate, as per the docs, since 0 means unlimited.
ref: http://jakarta.apache.org/commons/dbcp/configuration.html
> I think the "max_idle = 0" should be considered a bug
I agree. Want to open a JIRA issue and submit a patch for the config?
--- Noel
---------------------------------------------------------------------
To unsubscribe, e-mail: server-user-unsubscribe@james.apache.org
For additional commands, e-mail: server-user-help@james.apache.org
Re: query.bondedsender.org & 1000s of TIME_WAITs
Posted by ap...@bago.org.
> > netstat is still returning thousands of these lines:
> > tcp 0 0 10.235.221.9:46882 10.235.221.12:mysql TIME_WAIT
>
> Something is definitely not right, especially since JAMES
> will use a connection pool, and a very limited # of JDBC connections.
I think there is a problem with the default DBCP configuration of the
max_idle property.
I see that James by default set "max_idle = 0": I added
<max_idle>80</max_idle> to my datasource configuration and seems to work
better. (I have <max>80</max> in the same datasource configuration).
I think the "max_idle = 0" should be considered a bug (IIRC dbcp by default
use max_idle = 8, but james overwrite this with its default): can someone
confirm my diagnosis?
Stefano
---------------------------------------------------------------------
To unsubscribe, e-mail: server-user-unsubscribe@james.apache.org
For additional commands, e-mail: server-user-help@james.apache.org
RE: query.bondedsender.org & 1000s of TIME_WAITs
Posted by "Noel J. Bergman" <no...@devtech.com>.
Nathan,
I had not noticed it, but I can confirm that at least this morning:
$ dig @query.bondedsender.org 2.202.112.66.query.bondedsender.org
; <<>> DiG 9.2.1 <<>> @query.bondedsender.org
2.202.112.66.query.bondedsender.org
;; global options: printcmd
;; connection timed out; no servers could be reached
It seems that a new company has aquired
(http://www.returnpath.biz/aboutus/newsroom/press/release.php?id=31&flash=ye
s) that program, and perhaps they are having migration pains. The web site
(http://www.bondedsender.org/) is still up, and does not have any new
instructions. I'll ping the support folks there.
However, my server continues to work fine. It just can't whitelist people.
So I would suspect some other issue.
> netstat is still returning thousands of these lines:
> tcp 0 0 10.235.221.9:46882 10.235.221.12:mysql TIME_WAIT
Something is definitely not right, especially since JAMES will use a
connection pool, and a very limited # of JDBC connections.
--- Noel
-----Original Message-----
From: Nathan Cheng [mailto:nathan_w_cheng@yahoo.com]
Sent: Saturday, May 14, 2005 3:31
To: James Users List
Subject: query.bondedsender.org & 1000s of TIME_WAITs
Has anybody else noticed that query.bondedsender.org is apparently no longer
functioning? At least it is not responding
to ping, and apparently is the cause for James' SMTP service not working for
us the entire day: SMTP is working for the
first time today (technically yesterday--it's been a long "James" day) now
that I've commented out the following from
our James config file:
<!-- People on this list agree to pay a penalty if they send spam
<mailet match="InSpammerBlacklist=query.bondedsender.org"
class="ToProcessor">
<processor> transport </processor>
</mailet>
-->
I'm still not sure that this is the problem though, since netstat is still
returning thousands of these lines:
tcp 0 0 10.235.221.9:46882 10.235.221.12:mysql
TIME_WAIT
and it just seems like it's not right to have 3118 of these at the same
time, dying and being created seemingly by the
hundreds per second.
But SMTP is working now that bondedsender is no longer being consulted. But
maybe that's just because the load has died down and we still have an
unsolved problem some where.
Any ideas anyone?
Thanks,
Nathan
---------------------------------------------------------------------
To unsubscribe, e-mail: server-user-unsubscribe@james.apache.org
For additional commands, e-mail: server-user-help@james.apache.org