Re: cat /proc/net/tcp takes 0.5 seconds on x86_64

[Eric Dumazet <dada1@cosmosbay.com>]

David Miller a écrit :
> From: Eric Dumazet <dada1@cosmosbay.com>
> Date: Thu, 28 Aug 2008 09:13:13 +0200
> 
>> David Miller a écrit :
>>> We could not define a reasonable way to trigger hash table growth.
>>> GC attempts to keep a resident set of entries in the cache, and these
>>> heuristics are guided by the table size itself.  So if you grow the
>>> table too aggressively this never has a chance to work.
>> Maybe because of overcomplicated algos in net/ipv4/route.c, and
>> mixing "number of entries in cache", and "hash table size" things...
> 
> But it is exactly what we want to use to create equilabrium
> when the size is large enough.
> 
>> Fact is that nobody wants to have eight elements per hash bucket,
>> especially in case of DDOS.
> 
> Of this there is no doubt.  But also, nobody wants to use 1GB hash
> table simply because of poorly implemented interactions between GC and
> hash table growth :-)

Sure, but we have /proc/sys/net/ipv4/route/max_size, so hash table size would be
at most (sizeof(void *) * max_size) of course.

> 
> It was a series of patches I started posting in August 2006:
> 
> 	http://marc.info/?l=linux-netdev&m=115510979129323&w=2
> 
> There was a bunch of discussion, the locking and synchronization
> were worked out but the resize logic was not.

Thanks David, I remembered this stuff but was unable to find it again in my archives.

Give me some days/weeks to work on this subject.

(Hans removed from CC)