Watching performance on a DHCP Server

[Blake Hudson]

-------- Original Message  --------
Subject: Re: Watching performance on a DHCP Server
From: Blake Hudson <blake at ispn.net>
To: dhcp-users at isc.org
Date: Thursday, February 07, 2008 3:37:16 PM
>>
>> BTW, what would you say would be typical of DHCP performance for a 
>> machine that handles 250k IP addresses
>>
>> Does 130 4-way handshakes or 600 2-way handshakes per second sound 
>> right to you? (Based on the dhcperf tool)
>>
>> Larry Apolonio
>
>
> Larry, I'm interested in testing the performance of a DHCP server 
> recently installed for an ISP (~ 25k IP addresses) to ensure that 
> should something occur on the network that the DHCP server will not 
> become overloaded with requests.
>
> I had assumed that a DHCP server should be able to handle hundreds, if 
> not thousands of requests a second, given the simplicity of the 
> protocol. However, recent problems that others have experienced has 
> lead me to the conclusion that this may not necessarily be true.
>
> I looked for the dhcperf tool that you mentioned. However, it appears 
> the hosting site (http://www.nominum.com/testing_tools.php) no longer 
> offers this program and I am unable to find a copy elsewhere. Would 
> anyone happen to know where I can find a copy or know of any other 
> similar tools that offer the functionality to 'load test' a DHCP 
> server? - Sourceforge listed a few promising projects, however none 
> had any code/files available. I wouldn't mind a little scripting if 
> the underlying client/server handling is already done, again I want to 
> try to mimic (or exaggerate) a real world situation where all network 
> clients have lost their leases and are requesting new ones 
> simultaneously.
>
> Thanks in advance,
> -Blake
>
>
>
It appears the list is now maintained here: 
https://www.nominum.com/services/measurement_tools.php which link to the 
following FTP site: ftp://ftp.nominum.com/pub/nominum/dhcperf/1.0.1.0/


Unfortunately, my DHCP handling by a pretty beefy server looks mediocre 
-with results in the double digits.

By default in my distribution the leases file is stored in 
/var/lib/dhcpd/dhcpd.leases. This happens to be on a RAID1 array with 
15k scsi disks and iostat shows the array as being maxed out once it 
reaches ~ 300 I/O's per second. DHCP logging is done asynchronously to 
the same array (which normally experiences ~ 50 I/O ops). With CPU and 
memory barely breaking a sweat, this leads me to believe that the 
limitation is with the disks (lots of tiny writes).

I could move the leases file to a different array, or to tmpfs, but 
before I do I just want to know if these results are typical and that I 
have interpreted the test data correctly and made the correct 
determination as to the bottleneck.


Also, I would appreciate any anecdotal evidence with regards to how many 
requests are typical in a large network under normal (or abnormal) 
conditions. If 10,000 users all of a sudden came online, how many 
requests would they really generate per second?

Thanks in advance,
--Blake