[Snort-users] high packet loss - low throughput

waldo kitty wkitty42 at ...14940...
Fri Jul 19 12:37:49 EDT 2013


On 7/19/2013 05:16, Michal Purzynski wrote:
> So, anyone got some ideas how to debug and improve the situation? Or
> should I just assume that snort isn't capable of handling a per process
> 30Mbit - I can see a 5% packet loss now.

are you running a 64bit OS on those boxes or a 32bit one? which OS? you said 
(below) part of a security onion so i'm going to guess linux... now 64 or 32bit?

assuming *nix, what does top show?

   top -bn1 | head



> On 7/18/13 11:07 AM, Michal Purzynski wrote:
>> On 7/18/13 3:39 AM, waldo kitty wrote:
>>> On 7/17/2013 17:25, Michal Purzynski wrote:
>>>> On 7/17/13 11:01 PM, waldo kitty wrote:
>>>>> On 7/17/2013 16:04, Michal Purzynski wrote:
>>>>>> Hello,
>>>>>>
>>>>>> I can see a strange results on a local snort installation. Either I
>>>>>> don't understand something or the statistics aren't precise.
>>>>>> Please help
>>>>>> me understand.
>>>>>>
>>>>>> It's an (expanding) two hosts snort setup with 2 x E5-2620 0 @
>>>>>> 2.00GHz /
>>>>>> 64GB RAM each.
>>>>>> Intel x520 card.
>>>>>> Traffic is around 1Gbit to each host.
>>>>>> Around 3500 VRT only rules enabled.
>>>>>> 8 snort instances load balanced by the pf_ring.
>>>>> what else is this machine doing besides just snorting the traffic?
>>>> netsniff-ng, barnyard, snort and that's it. Part of a Security Onion,
>>>> but with most things (like Bro, argus, prads, etc) disabled.
>>>>>> The traffic loss is very high - up to 9% per instance (as reported by
>>>>>> Sguil which in turn read the snort logs and debug files). A single
>>>>>> instance gets from 90 - 150Mbits of traffic and from 10 - 20k pps. To
>>>>>> make it worse, the loss is not dependent on the traffic and/or pps at
>>>>>> all. Actualy, sometimes I get a 5% of loss on 50Mbits to a single
>>>>>> instance.
>>>>> what happens if you increase the number of snort instances which
>>>>> would thereby
>>>>> reduce the load on each of the instances?
>>>> I did it increasing from 6 to 8. And it won't help, really - if snort
>>>> cannot keep up with 50Mbit / instance stream...
>>> i'm not sure that it is snort, specifically... there is something
>>> causing the
>>> data to be flushed or lost before it has a chance to be processed...
>>> there are
>>> others running snort on pipes as large or larger...
>>>
>>> perhaps you are using protocol aware stream flushing and it needs
>>> tweaking?
>> Yes, it's enabled with the same settings. Reading about it and I don't
>> really want to disable it.
>>>
>>> ###############################################
>>> # Configure protocol aware flushing
>>> # For more information see README.stream5
>>> ###############################################
>>> config paf_max: 16000
>>>
>>>
>>> it may also be related to the timeout values in the stream5 settings??
>>>
>>>
>> No idea, that's why asking here :) Everything is default.



-- 
NOTE: No off-list assistance is given without prior approval.
       Please keep mailing list traffic on the list unless
       private contact is specifically requested and granted.




More information about the Snort-users mailing list