On Thu, 5 Nov 1998, LEE, Yui-wah wrote:

Hi Clement,

> Do you still have the SrvLog (in /vice/srv/) of the machine "devil" ?
> >From what you said, the createvol_rep script have correctly identified
> the two servers (argus and devil), but when it was trying to talk to
> devil, the RPC2 call failed (there was not even a binding).  The SrvLog
> may tell what went wrong.
The codasrv didn't run at that time but I haven't recognized it.
Now all works correct on "devil" but the problem has moved to "argus"

SrvLog tells:
12:04:53 New SrvLog started at Thu Nov  5 12:04:53 1998

12:04:53 Resource limit on data size are set to 2147483647

12:04:53 Server etext 0x818eed0, edata 0x820b388
12:04:53 RvmType is Rvm
12:04:53 Main process doing a LWP_Init()
12:04:53 Main thread just did a RVM_SET_THREAD_DATA

12:04:53 Setting Rvm Truncate threshhold to 5.

log_recover failed.
do_rvm_options failed
12:04:54 rvm_init failed RVM_EIO


It seems to be a problem within RVM.
I used rvmutl to solve:
rmvutl  open_log mylogfile 
        recover
        init_log
Then a restart from codasrv was necessary. Now the codasrv works fine.


Carsten

Carsten Ranfeld
Carsten.Ranfeld_at_informatik.tu-chemnitz.de
ESC :%s/windows/linux/g