Hi there,

Basically, if you want to set up two replicated servers, set up the
second one exactly like you set up the first one, except specify that
the system will not be a SCM, then point it at the SCM for your
network... looks like you've done that.

> I configured the replication server:

I think you've done everything fine. To make sure they're in sync, you
might try shutting down both servers, then bring up the SCM, then
bring up the first slave.

> Then I kept on going on the SCM:
>
> # cat /vice/db/VRList
> / 7f000000 1 01000001 0 0 0 0 0 0 0
> # volutil setlogparms 7f000000 reson 4
> V_BindToServer: binding to host clusty1
> Set Log parameters
> # volutil -h clusty2.mytest.de create_rep /vicepa /.1 7f000000
> V_BindToServer: binding to host clusty2.mytest.de
> RPC2_NewBinding to server clusty2.mytest.de failed with RPC2_NOBINDING (F)

You probably don't want to use the volutil program to create
replicated volumes (it is probably not the proper method anyway).
There is a tool called createvol_rep that will do the job very nicely.
Get on your SCM and, as root on the SCM, you would do this to make a
volume --

createvol_rep my.volumename my.first.server/vicepN my.second.server/vicepN ...

and so on. For example,

createvol_rep coda.test server1.mydomain.com/vicepa server2.mydomain.com/vicepb

Be sure to only run createvol_rep on the SCM, only while logged in as root.

I find that a lot of the subcommands of cfs and volutil simply do not
work, or do not work as you might expect. Generally, you shouldn't
really have to use volutil very much (I never do).

Does that work?

Regards, Sean
scaron_at_umich.edu