Coda File System

RE : Replication problem...

Date: Thu, 12 Aug 2004 08:48:13 +0200

> > # cat VRList
> > volROOT 7f000001 2 a000002 14000001 0 0 0 0 0 0 E0000103 volSHARE 
> > 7f000002 2 a000003 14000002 0 0 0 0 0 0 E0000103
> >  
> > # cat VSGDB
> > E0000100 servint1
> > E0000102 proxytest
> > E0000103 servint1 proxytest
> > I was just beginnig all my tests.
> > The 2 servers up. The first time I modified the file toto with the 
> > client on the non SCM server I could not see the modifications with 
> > the client on the SCM server. It sounds like suddently the 
> replication 
> > process failed
> Let you tell the contents of your volumes, where do you have 
> mountpoints and so on.
/coda/myrealm/ root volume volROOT, no files inside
/coda/myrealm/share mountpoint for volume volShare, just one file "toto"

> For a replication test you would not need another volume as 
> your rootvolume already is replicated. Having two volumes 
> just makes the picture harder to interpret.
Hmmm... I read it is harder to repair the root file than a mounted volume...
I have no specific needs... Just a simple system I can trust...
So what should be the correct configuration for a very simple replicated
file system : 
- just one entry in the VSGDB file and only one in the VRLIST file ?

> Have you got Coda working properly without replication?
I think so. Since I restarted the venus client everything look ok.

But I would like to understand why this problem occurs ?
I can not trust a system if it freeze without explanation...

> What is the path to you "toto" file?

> What does "cfs lv ." say in the directory where the file resides?
# cfs lv .
  Status of volume 0x7f000002 (2130706434) named "volSHARE"
  Volume type is ReadWrite
  Connection State is Connected
  Minimum quota is 0, maximum quota is unlimited
  Current blocks used are 3
  The partition has 378080 blocks available out of 956728
  Write-back is disabled

On the non-SCM server, the result is different on the following line :
  The partition has 378076 blocks available out of 956728

Is this normal ???

Thank you for the help.
Marc Cassuto.
Received on 2004-08-12 02:49:45