Coda File System

Re: repair issue

From: Jan Harkes <jaharkes_at_cs.cmu.edu>
Date: Fri, 2 Dec 2005 14:12:29 -0500
On Fri, Dec 02, 2005 at 11:56:58AM -0700, Patrick Walsh wrote:
> 	By and large our system is working great and we see few conflicts.
> Today, however, we got one and I am not sure how to fix it.
> 
> 	If I use the repair tool, I get this:
> 
> repair > beginrepair passwd_manage
> Too few directory entries
> Could not allocate replica list
> beginrepair failed.
> 
> 	If I use cfs beginrepair, the "global" file is a broken link.  And
> really, what can you do besides examine the conflict after you've used
> cfs to begin a repair?

It is a combination of a server-server (resolution) conflict and a
local-global (reintegration) conflict.

The reintegration probably failed because of the server-server conflict.
This is a problematic case because both types of conflicts are handled
very differently and the client can't fix the underlying server-server
conflict because it is 'hidden' by the reintegration.

Right now the options are, fix the server-server conflict from another
client which may then allow reintegration to succeed, or purge the cml
to get rid of the local-global conflict, and repair the remaining
server-server from this client.

> 	I can't rm it and don't know how to replace it or get rid of it.  Any
> suggestions?
    
cfs purgeml should get rid of the local-global (reintegration) conflict.

At this point you would only have the server-server conflict left, if
this is a file it can be completely removed with
    removeinc passwd_manage

or repaired with,
    cfs beginrepair passwd_manage
    cp passwd_manage/some_replica /tmp/passwd_manage.recovered
    cfs endrepair passwd_manage
    filerepair passwd_manage /tmp/passwd_manage.recovered

Jan
Received on 2005-12-02 14:13:18