Fw: Re: 2.6.12-rc2-mm2

Fw: Re: 2.6.12-rc2-mm2

[Andrew Morton]

Begin forwarded message:

Date: Fri, 08 Apr 2005 15:25:16 +0300
From: Indrek Kruusa <indrek.kruusa@tuleriit.ee>
To: Andrew Morton <akpm@osdl.org>
Cc: linux-kernel@vger.kernel.org
Subject: Re: 2.6.12-rc2-mm2


Andrew Morton wrote:

>ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.12-rc2/2.6.12-rc2-mm2/
>  
>

Wow... it responds despite the load average of 83.63 :)

http://www.tuleriit.ee/progs/img/pic1.png
http://www.tuleriit.ee/progs/img/pic2.png
http://www.tuleriit.ee/progs/img/pic3.png


dmesg is not clean though:

TCP: time wait bucket table overflow
TCP: time wait bucket table overflow
TCP: time wait bucket table overflow
printk: 392 messages suppressed.
TCP: time wait bucket table overflow
printk: 290 messages suppressed.
TCP: time wait bucket table overflow
TCP: time wait bucket table overflow
TCP: time wait bucket table overflow
TCP: time wait bucket table overflow
TCP: time wait bucket table overflow
TCP: time wait bucket table overflow
TCP: time wait bucket table overflow
TCP: time wait bucket table overflow
TCP: time wait bucket table overflow
TCP: time wait bucket table overflow
printk: 295 messages suppressed.
TCP: time wait bucket table overflow


What I did:
- "bombing" httpd with JMeter (from another computer)
- copying files from USB memory device to harddisk
- copying file with scp


Indrek

Re: Fw: Re: 2.6.12-rc2-mm2

[Herbert Xu]

Andrew Morton <akpm@osdl.org> wrote:
> 
> dmesg is not clean though:
> 
> TCP: time wait bucket table overflow
> TCP: time wait bucket table overflow
> TCP: time wait bucket table overflow
> printk: 392 messages suppressed.
> 
> What I did:
> - "bombing" httpd with JMeter (from another computer)

This is normal.  The number of TCP connections held by
your server in the TIME_WAIT state is exceeding
tcp_max_tw_buckets.

Check out tcp(7) on what this means and how it may or
may not be a problem for you.
-- 
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt

Re: Fw: Re: 2.6.12-rc2-mm2

[Indrek Kruusa]

Herbert Xu wrote:

>Andrew Morton <akpm@osdl.org> wrote:
>  
>
>>dmesg is not clean though:
>>
>>TCP: time wait bucket table overflow
>>TCP: time wait bucket table overflow
>>TCP: time wait bucket table overflow
>>printk: 392 messages suppressed.
>>
>>What I did:
>>- "bombing" httpd with JMeter (from another computer)
>>    
>>
>
>This is normal.  The number of TCP connections held by
>your server in the TIME_WAIT state is exceeding
>tcp_max_tw_buckets.
>
>Check out tcp(7) on what this means and how it may or
>may not be a problem for you.
>  
>

Thanks for your answer! I'll try to tune this parameter.
Indrek

Re: Fw: Re: 2.6.12-rc2-mm2

[Indrek Kruusa]

Herbert Xu wrote:

>Andrew Morton <akpm@osdl.org> wrote:
>  
>
>>dmesg is not clean though:
>>
>>TCP: time wait bucket table overflow
>>TCP: time wait bucket table overflow
>>TCP: time wait bucket table overflow
>>printk: 392 messages suppressed.
>>
>>What I did:
>>- "bombing" httpd with JMeter (from another computer)
>>    
>>
>
>This is normal.  The number of TCP connections held by
>your server in the TIME_WAIT state is exceeding
>tcp_max_tw_buckets.
>
>Check out tcp(7) on what this means and how it may or
>may not be a problem for you.
>  
>


This poor Celeron machine has only 128MB RAM and tcp_max_tw_buckets were 
calculated (ipv4/tcp.c) to be 16384. I have found
that increasing this number by 2048 (that's 12.5%) = 18432 is enough to 
not get this overflow warning again (load average was over 90). Some 
other facts:

- my current machine has 512MB RAM and calculated tcp_max_tw_buckets is 
180000 (!)
- from Documentation/networking/ip-sysctl.txt: This limit exists only to 
prevent simple DoS attacks


I believe that those values are somewhat miscalculated. At least by 
changing calculation for machines with low memory a bit you will get 
some defence against poor shouting students who had to build 
load-balancing web-clusters ;)

thanks,
Indrek

Re: Fw: Re: 2.6.12-rc2-mm2

[Indrek Kruusa]

Herbert Xu wrote:

>Andrew Morton <akpm@osdl.org> wrote:
>  
>
>>dmesg is not clean though:
>>
>>TCP: time wait bucket table overflow
>>TCP: time wait bucket table overflow
>>TCP: time wait bucket table overflow
>>printk: 392 messages suppressed.
>>
>>What I did:
>>- "bombing" httpd with JMeter (from another computer)
>>    
>>
>
>This is normal.  The number of TCP connections held by
>your server in the TIME_WAIT state is exceeding
>tcp_max_tw_buckets.
>
>Check out tcp(7) on what this means and how it may or
>may not be a problem for you.
>  
>
Wait a moment.... with kernel 2.6.10-oci7 (Sam) load average is 3-4x 
lower and system does not reach this tcp_max_tw_buckets limit. If it is 
desired I can do more tests with different kernels.

Indrek