LVS
lvs-users
Google
 
Web LinuxVirtualServer.org

Re: Really strange least connection problem

To: lvs-users@xxxxxxxxxxxxxxxxxxxxxx
Subject: Re: Really strange least connection problem
From: Jason Steenblik <jsteenblik@xxxxxxxxxx>
Date: Mon, 9 Apr 2001 16:07:32 -0700
Thanks,

The patch is on and working great.  I only have a half a day's worth of
testing on it but so far so good. Here is ipvsadm -ln now.  

IP Virtual Server version 0.2.8 (size=4096)                    
Prot LocalAddress:Port Scheduler Flags                         
  -> RemoteAddress:Port               Forward Weight ActiveConn
InActConn
TCP  172.19.0.5:25 lc
  -> 172.18.2.16:25                   Masq    1      1          205       
  -> 172.18.2.15:25                   Masq    1      1          193       
  -> 172.18.2.14:25                   Masq    1      1          183       
  -> 172.18.2.13:25                   Masq    1      1          206       
  -> 172.18.2.12:25                   Masq    1      0          207       
  -> 172.18.2.11:25                   Masq    1      0          223       
  -> 172.18.2.10:25                   Masq    1      0          223       
  -> 172.18.2.9:25                    Masq    1      0          227       
  -> 172.18.2.8:25                    Masq    1      1          184       
  -> 172.18.2.7:25                    Masq    1      1          203       
  -> 172.18.2.6:25                    Masq    1      1          202       
  -> 172.18.2.5:25                    Masq    1      1          165       
  -> 172.18.2.4:25                    Masq    1      2          155       
  -> 172.18.2.3:25                    Masq    1      0          223       
  -> 172.18.2.2:25                    Masq    1      0          197       
  -> 172.18.2.1:25                    Masq    1      3          164       

Thank You,
Jason


On Fri, Apr 06, 2001 at 11:16:16PM +0800, Wensong Zhang wrote:
> 
> 
> Hi Jason,
> 
> Thanks for the info. It is mostly some race in state transition on SMP
> box, which leads to those negative problem.
> 
> We added a simple spinlock to fix this problem, since the collision seldom
> happens, there won't be performance drawback. Please apply the attached
> patch to ipvs-0.2.8 tar ball, build and test it on your smp box.
> 
> Thanks,
> 
> Wensong


<Prev in Thread] Current Thread [Next in Thread>