Re: Problems with takeover even though master is online

Hi agian,

nope.. haven't tried -v yet .. here is my lvs.cf. Its the same on both 
machines and yes, they can ping themselves over eth0 (public 
interface). I had the timeout at 8 seconds . not upped it to 15 
seconds. The problem has not surfaced again so far.

service = lvs
primary = <IP1>
rsh_command = ssh
backup_active = 1
backup = <IP2>
heartbeat = 1
heartbeat_port = 1050
keepalive = 6
deadtime = 15
reservation_conflict_action = preempt
debug_level = MAX

network = nat
nat_router = 192.168.1.254 eth1:1
nat_nmask = 255.255.255.0
virtual Web_Server {
      active = 1
      address = <IP3> eth0:1
      port = http
      persistent = 1
      pmask = 255.255.255.255
      send = "GET / HTTP/1.0\r\n\r\n"
      expect = "HTTP"
      load_monitor = rup
      scheduler = wlc
      protocol = tcp
      timeout = 5
      reentry = 10

      server web1 {
          address = 192.168.1.10
          active = 1
          weight = 1
      }
      server web2 {
          address = 192.168.1.11
          active = 1
          weight = 1
      }
}

virtual CMSTC {
      active = 1
      address = <IP4> eth0:2
      port = 8080
      pmask = 255.255.255.255
      load_monitor = rup
      scheduler = lc
      protocol = tcp
      persistent = 0
      timeout = 6
      reentry = 15
      server cms {
          address = 192.168.1.40
          active = 1
          weight = 1
      }
}

virtual CMSPG {
      active = 1
      address = <IP4> eth0:2
      port = 5432
      pmask = 255.255.255.255
      load_monitor = rup
      scheduler = lc
      protocol = tcp
      persistent = 0
      timeout = 6
      reentry = 15
      server cms {
          address = 192.168.1.40
          active = 1
          weight = 1
      }
}

I also have to add that I'm running this system now for more than a 
year and never encountered the problem before. Therefor I don't know if 
I'll be able to reproduce the problem.

On Freitag, Okt 18, 2002, at 17:45 Europe/Vienna, Mike McLean wrote:

> Have you tried added -v to the pulse command line (in 
> /etc/init.d/pulse)?  This will cause debugging messages to be written 
> to the syslog, it may help to diagnose the problem.
>
> It looks like everything was fine for two hours.  Hmm, this seems like 
> some sort of timeout.  Is there anything at all unusual about your 
> network layout?
>
> I'd be curious to see the results of a tcpdump on port 539 (or 
> whatever port you're heartbeating on) at two different times:
>   - during the initial "working" period
>   - after the backup takes over
>
> Also, could you post your lvs.cf?
>
----
   Dipl.-Ing. Peter Burgstaller
   Technical Assistant and System Administrator
   @ all information network & services gmbh
   email: [email protected]
   phone: +43 662 452335
   fax  : +43 662 452335 90