4.5, and dbmail is from (slightly old) CVS.
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61) 0500 83 82 81
tions fail
because the max connections is exceeded. I am not at
all sure of the sequence of events, but this seems right.
Bye for now,
Philip
regards roel
- Original Message -
From: "Philip Warner" <[EMAIL PROTECTED]>
To:
Sent: Sunday, September 22, 2002 10:4
site. Ordinarily this would also be fine, except I
am not seeing *any* messages from most of these processes again. Is that
normal?
Philip Warner| __---_
Albatross Consulting Pty
r callaghan
logging out [message=0, octets=0]
Sep 25 00:01:51 mag dbmail/pop3d[3463]: Maximum # of connections reached,
committing suicide...
--------
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |---
list of pids looks like.
--------
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (
this is
a good idea, please add it ti the ToDo under my name...
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel
this size difference wasn't so big.
Are you running a nightly vacuum process?
----
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 00
ages is something like:
From [EMAIL PROTECTED] Tue Oct 15 11:16:37 2002
Received: from ...etc
Message-ID: <[EMAIL PROTECTED]>
Date: Tue, 15 Oct 2002 11:16:10 +1000
To: [EMAIL PROTECTED]
From: Philip Warner <[EMAIL PROTECTED]>
Subject: another test
Mime-Versi
?
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61) 0500 83 82 81 | _ \
Fax: (+61) 0500 83 82 82
mime.c problems I reported?
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61) 0500 83 82 81
his to the list (or at least Roel/Eelco) before.
There are also patches to use PostgreSQL with postfix at:
http://www.rhyme.com.au/postfix.html
--------
Philip Warner| __---_
Albatross Consultin
looked closely at bounce.c to see what else we might want to
substitute.
--------
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) _
.
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61) 0500 83 82 81 | _ \
Fax: (+61) 03 5330 3172
h to fix
the mime parsing bug).
----
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61) 0500 83 8
?
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61) 0500 83 82 81
which sometimes causes a segfault when reading large files. I notice that
the CVS version of pipe.c has been fixed to read:
...strblock = (char *)my_malloc(READ_BLOCK_SIZE + 1)...
but not forward.c.
-------
FROM messageblks WHERE message_idnr = ::bigint
ORDER BY messageblk_idnr;
Any chance of letting me know the PG version, and doing and EXPLAIN on
these? I do not see this behaviour.
----
Philip Wa
idea to remove the LIMIT 1 in general since it
will cause the backend process to retrieve the entire message (not good for
big messages). But it is obviously necessary to avoid a sequential scan of
the messageblks table!
--------
ning?
----
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61) 0500 83 82 81 | _ \
Fax: (+61) 03 5330 3172 | ___ |
if you are not doing it already, is a
regular vacuum of the entire database, and probably a *very* regular vacuum
of frequently UPDATED tables (eg. we vacuum the 'users' table every 5 minutes).
------
this case].
----
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61) 0500 83 82 81 | _ \
Fax: (+61) 03
tegy as you on an empty DB; I
think you should run an ANALYZE then try the query again. We run an analyze
every night.
--------
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N.
At 01:05 AM 22/11/2002 +1100, Philip Warner wrote:
I have confirmed that I get the same strategy as you on an empty DB; I
think you should run an ANALYZE then try the query again. We run an
analyze every night.
I should point out that as of recent versions of PG (7.1?), ANALYZE is a
lot
costs to estimate which plan is really the cheapest
The planner does know that indexes can be used to return data in sorted
order, so choosing to use the messageblk_idnr index is not a bad strategy
(for small amounts of data).
----
Phil
would be quite willing to put a non-trivial wager on the fact that an
ANALYZE will result in your DB working as you would hope!
--------
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B
At 03:30 PM 21/11/2002 +0100, Eelco van Beek - IC&S wrote:
I'm just curious what the pgsql mailinglist will reply.
No surprises there:
(You have done a vacuum analyze recently, I trust...)
----
Phil
o part of the pgsql mailing list request)
--------
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61) 0500 83 82 81 |
ike asking a pilot to fly a plane with no radar and a blackened
windscreen: "Just do what you did last time".
This is especially true if the database data profile has changed substantially.
------
far as I can tell. As a result it sends "-1" formatted as an unsigned long
long as the message length, and dies when it tries to read the buffer.
Any help (or patches!) would be greatly appreciated. Perhaps this is a
problem with the storage system, or even the mail tha
more likely to
be terminated with a NULL.
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61) 0500 83 82 81
At 04:59 PM 29/11/2002 +, Sam Przyswa wrote:
I got it but on which version or CVS date is it for ?
I have used it on RC3 and it was built against the current tip.
Philip Warner| __---_
Albatross
pted
indices. The indices aren't corrupt, the database is just too big!
Again, this is almost certainly tuning parameters. I had the same problem
(growth of about 140MB per day), until I fixed some settings.
--------
Philip Warn
will decrease the FSM size and/or
reduce the vacuum frequency.
I hope this helps.
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /
At 05:16 PM 3/12/2002 +1100, Philip Warner wrote:
set max_fsm_pages to 4
vacuum every two hours
vacuum 'users' every 5 min
analyze every hour
dbmail-maintenance -p -d -f daily.
I should have mentioned two things:
(a) with the above settings, the server load is minimal (<1),
At 02:19 AM 8/12/2002 -0300, A C Mirand wrote:
Where I can Start checking?
You need to start dbmail-pop3d (and dbmail-imapd) manually.
Philip Warner| __---_
Albatross Consulting Pty. Ltd
At 02:33 AM 8/12/2002 -0300, A C Mirand wrote:
Both are already running..
Can you telnet to port 110?
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498
At 02:19 AM 8/12/2002 -0300, A C Mirand wrote:
I can't read the emails by imap or pop3d
So what did this actually mean? Do you have error messages from a client
running on local host?
Philip W
quite deep, and a user
may have many, each vacation entry must also have a 'from-address' field to
allow a reasonable 'from' field to be set.
Any chance of some/all of these patches being applied some time?
--------
dvantages of a *real* database
Hope this helps
--------
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) _
At 11:40 AM 18/12/2002 +1100, Philip Warner wrote:
One of the features that is often overlooked when codsidering DBMail is
that you get all the advantages of a *real* database
Sorry...send was hit too soon...
One of the features that is often overlooked when considering DBMail is
that you
At 02:00 PM 18/12/2002 +1300, Craig Whitmore wrote:
The postgreSQL
version hasn't been changed as I don't know if if it has a similar command
as REPLACE that mysql does.
It doesn't, so no need to worry about that.
-------
tive usage pattern
(dbmail-maintenance, day/nigh etc). Once you have that, I can give more
specific advice about the best settings (or you can use the link above to
work this out for yourself).
--------
ings to use to at least stop the growth.
--------
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61) 0500 83 82 81 |
on of PG are you running?
--------
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61) 0500 83 82 81 | _ \
hat VACUUMing a cleaner database will put less
strain on your server.
--------
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---
to a much lower value.
Note that to set the MAX_FSM_* parameters you need to restart the
postmaster; so it might be good to do it all together.
----
Philip Warner| __---_
Albatross Consulting Pty.
also arrange downtime to do a vacuum FULL
followed by hourly VACUUMs? You may find that the system will not die if
vacuum is run more regularly.
----
Philip Warner| __---_
Albatross Consul
At 02:07 PM 18/12/2002 +0100, Roel Rozendaal - IC&S wrote:
postgresql is version 7.2.
There have also been several improvements & bug fixes in PG since 7.2; you
might consider upgrading to 7.3 or 7.3.1.
----
Phili
y 24 hours), and I will look into it. Make sure
that you capture the VACUUM VERBOSE output from just after you run
dbmail-maintenance etc.
I am on holiday at the moment, so don't expect a fast turn-around...
----
Phil
ame from
pg_class' (SQL).
enjoy your holiday :)
Good so far...
--------
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __
.
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61) 0500 83 82 81 | _ \
Fax: (+61) 03 5330 3172
good.
Thanks for the help.
--------
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61
ackup.
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61) 0500 83 82 81 | _ \
Fax: (+61) 03 5330 3172 | ___ |
.
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61) 0500 83 82 81 | _ \
Fax: (+61) 03 5330 3172 | ___ |
Http
recent CVS version. And it
allows the specification of from/to fields on the vacation field. It is
just a modified version of the patch submitted by Moisio Juha modified
somewhat.
--------
Phi
y sensible, which Eudora does not seem to do)?
regards roel
On Wed, 2002-07-10 at 06:28, Philip Warner wrote:
>
> Not sure if this is limited to my installation or not, but I would be
> interested to know if anyone else sees this (eg. with mySQL). When I
> 'rebuild' an IMA
auses.
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61) 0500 83 82 81 | _ \
Fax:
Further investigations of my undead-pop-message problem has revealed
somthing that confuses me: the unique_id field in the messages table is not
unique.
Should they be?
Philip Warner
At 18:05 6/08/2002 +1000, Philip Warner wrote:
Further investigations of my undead-pop-message problem has revealed
somthing that confuses me: the unique_id field in the messages table is
not unique.
Should they be?
Looking at the code in pipe.c, it uses the current time to build the ID
ed (eg. as long as it is not forked), we
could just append a sequence number.
But, AFAICT, the current code does not produce unique IDs.
Philip Warner wrote:
At 18:05 6/08/2002 +1000, Philip Warner wrote:
Looking at the code in pipe.c, it uses the current time to build the ID;
several mes
.
Thanks!
Do you think this is likely to cause a POP client to fail to mark some
messages as 'seen'? (my original undead-pop-message problem)
--------
Philip Warner| __---_
Albatross Consulting Pty. Lt
since it only uses the current time.
--------
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61) 0500 83 82 81 | __
behaviour'; I think you are
right, but if the 'message-id' passed to create_unique_id is not unique, we
will have a problem.
--------
Philip Warner| __---_
Albatross Consulting Pty
much.
I just appended a '.' followed by the value of message_idnr.
--------
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 0
__
Dbmail mailing list
Dbmail@dbmail.org
https://mailman.fastxs.nl/mailman/listinfo/dbmail
----
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 7
h distinct destination user; this is
basically to allow spamassassin to run and apply per-user settings.
--------
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 0
There is a minor problem with the CVS version of the pgsql files; they
currently have a full path for pgsql headers, which may conflict with the
location specified when using 'configure'. Patch attached.
--------
Phi
At 11:10 10/08/2002 +0200, Eelco van Beek - IC&S wrote:
e-mail alias
[EMAIL PROTECTED] !/bin/spamassasin|/usr/local/sbin/dbmail-smtp -u eelco
Thanks; it's exactly what we need, especially when combined with spamd/spamc:
[EMAIL PROTECTED] !/blah/spamc -f -u
At 11:10 10/08/2002 +0200, Eelco van Beek - IC&S wrote:
[EMAIL PROTECTED] !/bin/spamassasin|/usr/local/sbin/dbmail-smtp -u eelco
I'm using the above to insert a filter before delivery, but lines starting
with 'From' (notably the first line of the message) are being escaped
(prepending a
At 00:00 11/08/2002 +1000, Philip Warner wrote:
At 11:10 10/08/2002 +0200, Eelco van Beek - IC&S wrote:
[EMAIL PROTECTED] !/bin/spamassasin|/usr/local/sbin/dbmail-smtp -u
eelco
I'm using the above to insert a filter before delivery, but lines starting
with 'From'
do everything DBMail can in
this regard (or so I believe?).
Further, if we want to add more filters in the future, or change the
filter, we just need to update the view.
Philip Warner| __---_
Al
zarre mail configurations.
--------
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61) 0500 83 82 81 | _ \
Fax: (+61) 0500 83 82 82 |
CVS version of pipe.c has been fixed to read:
...strblock = (char *)my_malloc(READ_BLOCK_SIZE + 1)...
but not forward.c.
--------
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.
CVS version of pipe.c has been fixed to read:
...strblock = (char *)my_malloc(READ_BLOCK_SIZE + 1)...
but not forward.c.
--------
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.
Here's a patch that fixes the buffer overrun in forward.c that I mentioned
earlier.
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 65
ried about yesterday.
___
Dbmail mailing list
Dbmail@dbmail.org
https://mailman.fastxs.nl/mailman/listinfo/dbmail
--------
Philip Warner| __---_
Albatross Consultin
ship with concomitant rights over a user).
----
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61) 0500
We would like to log failed login attempts. Would other people be
interested in this (optional) feature, and perhaps more importantly, would
such patches be likely to be included in the main distribution?
Philip Warner
have the whole process bundled.
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61) 0500 83 82 81
a request for those as well.
Hope we will finish this job soon together & conquer the world with dbmail..
regards roel
--------
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.
not deliver or retrieve (ie. fully disabled)
--------
Philip Warner| __---_
Albatross Consulting Pty. Ltd. |/ - \
(A.B.N. 75 008 659 498) | /(@) __---_
Tel: (+61)
81 matches
Mail list logo
