[zeromq-dev] Fwd: Exact matching on subscription topics

[Andrew Hume]

good but expected work.
i reimplemented the DFA just in time stuff (that alfie did for grep) for the 8th edition unix,
and there are better techniques for depleting the cache.
but in any case, the DFA can be exponentially big, and you see that more often
than you would think.

On Jan 18, 2012, at 7:43 PM, Martin Sustrik wrote:

> RabbitMQ guys have done some actual performance testing in this area. You may want to have a look at what they've found out:
> 
> http://www.rabbitmq.com/blog/2010/09/14/very-fast-and-scalable-topic-routing-part-1/
> 
> http://www.rabbitmq.com/blog/2011/03/28/very-fast-and-scalable-topic-routing-part-2/
> 
> Martin
> 
> On 01/19/2012 11:34 AM, John D. Mitchell wrote:
>> On Jan 18, 2012, at 18:25 , Martin Sustrik wrote:
>> [...]
>>>> as for regex, be careful. regex performance is NOT wire speed EXCEPT
>>>> for the case of substring and in many case, multiple substrings
>>>> (the unix fgrep case). plus, relatively few folk are adept with complex
>>>> regex expressions.
>>> 
>>> Ack.
>> 
>> FYI, check out the Thompson-DFA based regex vs. the potentially exponential slowdown of e.g. backtracking-based regex engines. E.g.:
>> http://swtch.com/~rsc/regexp/regexp1.html
>> 
>> Hope this helps,
>> John
>> 
>> _______________________________________________
>> zeromq-dev mailing list
>> zeromq-dev at lists.zeromq.org
>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
> 


------------------
Andrew Hume  (best -> Telework) +1 623-551-2845
andrew at research.att.com  (Work) +1 973-236-2014
AT&T Labs - Research; member of USENIX and LOPSA




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.zeromq.org/pipermail/zeromq-dev/attachments/20120118/efdd068d/attachment.htm>