El mar, 09-10-2007 a las 02:20 +0100, Andrew Doran escribió:

When 30's is online and on local dynamic dns pages is when it happend
that the lan network is worried. The next one that this will be (30's
online) I post the command that now is 

netstat -b 1 
    ex0 in             ex0 out          total in           total
out      
     bytes               bytes             bytes
bytes      
3486357895           774457742        3489923187
778023034      
       198                 178               198
178      
        66                 178                66
178      
       425                 349               916
840      
        66                 178                66
178      
        66                 178                66
178      
        66                 178                66
178      
        66                 178                66
178      
        66                 178                66
178      
        66                 178                66
178      
       164                 178               164
178      
      1080                1302              1080
1302      
     21094               22097             21094               22097

sysctl -a | grep mbuf
kern.mbuf.msize = 256
kern.mbuf.mclbytes = 2048
kern.mbuf.nmbclusters = 1024
kern.mbuf.mblowat = 16
kern.mbuf.mcllowat = 8


thank's and regards


> On Sat, Oct 06, 2007 at 08:54:43AM +0200, agtdino wrote:
> 
> > I'm getting general slow network performance in NetBSD vs Centos, with
> > the same Hardware,  ( PIV- 500MB RAM 3COM 10/100Mb ) RAID0 for software
> > with CCD -Concatenated disk device- ) 
> > 
> >  Services in the server : 
> > 
> >  - Squid cache 
> >  - Dns only Caching Name Server
> > 
> > For client are 30 machines all with the same hardware ( PIV - 256MB RAM
> > 3com 10/100Mb ) with windows 2000
> 
> It should not be a problem serving 30 active users with this machine.
>  
> > There is a guide to provide high extreme performance with netbsd and
> > squid?
> 
> Have you isolated the cause of the slowness? "netstat -b 1" will show you
> network traffic. Have a look at the interfaces with "netstat -i" and see if
> there are errors. Run "systat vmstat" look at the area describing disk I/O
> to see if any of the disks are under pressure. Also have a look at
> PAGING/SWAPPING in the top right to see if the system is running low on
> memory.
> 
> Thanks,
> Andrew