; a look on restrictions classes and postfix' support for MySQL table lookups.
>
>
> The main advance of a network-backend-less installation is: less points of
> failure.
Or use a policy server that caters for MySQL being down, therefore no point
of failure.
Cami
--
as by removing unneeded indexes before
the purge and re-creating them after. I don't know why Cami
sees deletes in msgrcpt based on time_num to be significantly faster.
I'm assuming that Mark has benchmarked the difference in
speed between the time_num index approach vs the DELETE
CASCADE ap
Paolo Cravero as2594 wrote:
Cami wrote:
Keep far far away from these queries, they are slow.
But probably keep data consistent...
They do the same thing really.
DELETE FROM quarantine WHERE time_num < UNIX_TIMESTAMP() - 7*24*60*60;
(make sure you have an index on time_num)
From h
60*60;
DELETE FROM msgs WHERE time_num < UNIX_TIMESTAMP() - 60*60 AND content IS NULL;
DELETE FROM quarantine WHERE time_num < UNIX_TIMESTAMP() - 7*24*60*60;
(make sure you have an index on time_num)
Issuing an "EXPLAIN" on those queries th
backup SQL
servers which you can write the quarantined message
into, unfortunately it doesn't mention which was
used when quarantining the message in question.
Cami
---
This SF.net email is sponsored by: Splunk Inc. Do you grep through log fi
5): 250 Ok: queued as 9B84782285
Passed, <> -> <[EMAIL PROTECTED]>, quarantine sVD9s2SBe3Q1, Message-ID: ,
Hits: -
TIMING [total 1400 ms] - [...] AV-scan-1: 194 (14%)38, AV-scan-2: 775
(55%)93, [...]
Both doesn't detect it.
Have you configured ClamAV to detect it?
set DetectB
s able to detect broken achieves and
broken executables.
Cami
---
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy a
spam
message? (If there is trouble with the database when
quarantining, perhaps there should be a "counter" to keep
track of errors. Ie, should amavisd-new get more than 5
database connect() or MySQL INSERT failures in the period
of N minutes, it should switch
rge system with millions of quarantined messages
all the table scans are taking hours and hours to
complete a simple cleanup.
Comments?
Cami
---
SF.Net email is sponsored by:
Tame your development challenges with Apache's Geronimo App Server
Cami wrote:
Mark Martinec wrote:
If this is not the case, it would be worth taking a look at level 4 or 5
log and see how the sender address lookups are being done.
Attached is a legit case of where this is happening.
[EMAIL PROTECTED] is the recipient, [EMAIL PROTECTED]
is the sender which
whitelisted in the database.
Cami
Oct 7 18:11:10 spamwall01.mweb.co.za amavis[3478]: (03478-01-74)
check_mail_begin_task: task_count=74
Oct 7 18:11:10 spamwall01.mweb.co.za amavis[3478]: (03478-01-74) lookup
(debug_sender) => undef, "[EMAIL PROTECTED]" does not match
Oct 7 18:11:
Mark Martinec wrote:
Ok, lets see a specific case.
Ok, this guy for example is suppose to be opted out completely.
(Log attached, and for the life of me i don't see any mention
about queries against mysql)
Cami
Oct 3 18:14:16 spamwall01.mweb.co.za amavis[11653]: (11653
idn't have SQL lookups
enabled.
Sorry for wasting your time.
Cami
---
This SF.Net email is sponsored by:
Power Architecture Resource Center: Free content, downloads, discussions,
and more. http://solutions.newsforge.com/ib
Mark Martinec wrote:
Cami,
<[EMAIL PROTECTED]>,<[EMAIL PROTECTED]>,Blocked,Hits=7.79,
^^ Broken..
<[EMAIL PROTECTED]>,<[EMAIL PROTECTED]>,Passed,Hits=-
^^ Working..
Currently all the machines part of amavisd-new serverfarms
are the same software configuration/ver
8:51:10 spamwall04.mweb.co.za amavis[23746]: (23746-01-84)
<[EMAIL PROTECTED]>,<[EMAIL PROTECTED]>,Blocked,Hits=7.6,Message-ID=<[EMAIL PROTECTED]>,Size=2556
Here it appears that whitelisting is broken again.
Currently all the machines part of amavisd-new serverfarms
are the same so
ay
or malicious clients, or if someone insists on using amavisd-new in a
pre-queue filtering setup; suggested by Tuomo Soini;
How does amavisd-new handle pre-queue filtering if a message is destined
to 2 different people with 2 different message size limits?
ie, user1 = 10M, user2 = 20M and th
MJ wrote:
Cami wrote:
5.0 is a bit low for an ISP, especially if you are using all
the additional rules provided by http://www.rulesemporium.com.
That being said, i recommend that you collect statistics on the
amount of mail that score between 5.0 -> 6.3 so you can see the
type of impact
the
type of impact the change will have, and try to guestimate the
amount of false positives doing so will cause.
Cami
---
SF.Net email is Sponsored by the Better Software Conference & EXPO
September 19-22, 2005 * San Francisco, CA * Deve
x27;m misunderstanding, how is there no practical use?
Existing policy servers can be enabled/disabled on a per-recipient
basis.
Cami
---
SF.Net email is Sponsored by the Better Software Conference & EXPO
September 19-22, 2005 * San Francisco, CA * De
best choice
is to let mail with bad headers pass, perhaps
with $defang_bad_header=1;
Perfect, thanks..
Cami
---
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward
headers?
We have a number of users hopping up and down
about mails from senders already whitelisted
and mails are ofcourse not making it through.
Comments?
Cami
---
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM
Hi Mark,
Cami wrote:
X-Amavis-Alert: BAD HEADER MIME error: error: part did not end with
expected boundary
Jun 6 08:24:42 spamwall03.mweb.co.za amavis[7005]: (07005-03-12) BAD
HEADER from <>: MIME error: error: part did not end with expected boundary
Jun 6 08:24:42 spamwall03.mweb
]: (07005-03-12)
<>,<[EMAIL PROTECTED]>,Passed,Hits=-3.3,Message-ID=<[EMAIL PROTECTED]>,Size=51458
What am i missing? I'm using amavisd-new-2.3.1.
Regards,
Cami
---
This SF.Net email is sponsored by: NEC IT Guy Games. How
stfix pre-queueing mechansim. (the pre-queue
tends not to scale for very large sites.)
Cami
---
This SF.Net email is sponsored by Oracle Space Sweepstakes
Want to be the first software developer in space?
Enter now for the Oracle Space Sweepstakes!
24 matches
Mail list logo
