On Fri, 18 Nov 2011, Robert Bandl wrote:

> However, if you're able to run with asynchronous resolver like c-ares,
> please note that while c-ares does perform async resolve which should
> greatly improve speed when using single thread and multi inferface, it is
> (Daniel isn't gonna like this comment :)) a fairly slow library. In fact
> it's so slow that it'd most likely take longer than 1.75 secs to download
> 1500 urls if you create an easy handle for each url and add it to multi
> interface.

Actually I _do_ like the comment since it brought something to my attention I
wasn't aware of.

I won't object to the numbers since I haven't run your tests. But... I never
heard anyone before complain on c-ares speeds and c-ares really doesn't do a
lot and I find it hard to imagine that it alone can make for any significant
time loss. I've not seen anything like that during my tests - although it was
a while since I did any.

I think that if there's a noticable speed difference for a single handle with
c-ares compared to any of the other resolver backends, there's either a flaw
in your test or there's a bug in either libcurl or c-ares.

Additionally, if you're ramping up the number of parallel connections to 1K
or why not 10K or 40K the c-ares way will keep your memory usage to sane
levels while using a thread for each connection will not.

-- 
  / daniel.haxx.se
-------------------------------------------------------------------
List admin: http://cool.haxx.se/list/listinfo/curl-library
Etiquette:  http://curl.haxx.se/mail/etiquette.html