(Illustration by Gaich Muramatsu)
On Wed, Jul 18, 2001 at 05:52:13AM +0800, tin wrote: > thank you for your reply jan > > i have somethings still confuse > > 1. E0000100 mean server group ID? Yes, it identifies a group of Coda servers (the volume storage group). The server ID's are in /vice/db/servers. > 2. If I assume E0000100 is server ID ,then how to add non-SCM server in > E0000100 after doing createvol rootvolume E0000100 /vicepa ? It is at the moment not possible to grow or shrink a volume storage group once it has any volumes associated with it. > 3. E0000100 cannot replicate with E0002001 right? Nope. E0000100 is in a standard setup the group that contains a single server, the SCM. In my previous example the E0002001 group was set up with 2 servers, the SCM and the new server. If you want to have a rootvolume in the E0002001 group once the second server is up and running. Create a new volume on the new VSG, change /vice/db/ROOTVOLUME to point to the new volume. It is necessary to reinitialize all clients so that they can pick up the fact that they have to start using the new rootvolume. Then it is just a matter of recreating the mountpoints and copying files from the old rootvolume. SCM# createvol_rep rep:rootvolume E0002001 SCM# echo rep:rootvolume > /vice/db/ROOTVOLUME client# /etc/rc.d/init.d/venus.init stop client# touch /usr/coda/venus.cache/INIT client# /etc/rc.d/init.d/venus.init start client$ cd /coda client$ cfs mkm /coda/oldroot rootvolume client$ cfs mkm usr usrvolume client$ cfs mkm home homevolume JanReceived on 2001-07-17 18:08:26