> On Mon, Nov 08, 2004 at 09:01:42AM -0800, redirecting decoy wrote:
>> I ran into the same problem as you, where I could not
>> delete volumes.  In order to get it working I did the
>> following:
>>
>> edit purgevol_rep:
>> right after line 45, add "$REPVOLNAME=$2", right
>> before "dryrun=0"
>> so it should look like this:
>> REPVOLNAME=$1
>> if [ "$1" = "--kill" ]
>> then
>>     REPVOLNAME=$2               #MINE
>>     dryrun=0
>>     shift
>> else
>>     dryrun=1
>>     echo "Only testing, use 'purgevol_rep --kill $1'
>> to really purge the volume"
>> fi
>
> Good catch, it looks like I moved the REPVOLNAME= up too far when
> removing the VSGDB dependencies. Normally the 'shift' drop the --kill
> argument and after this 'if/else/fi' statement $1 will always contain
> REPVOLNAME.
>
I had done so already, so that wasn't really the problem.

>> Also, change line 55 to:
>> SCM=`cat /vice/db/scm`
>> volinfo=`${exec_prefix}/bin/getvolinfo $SCM $REPVOLNAME`
>
> That shouldn't be necessary... Unless ofcourse you have bound your
> server to a specific IP-address. Hmm, that does create a bit of a
> problem, what if we have several servers running on the same machine
> (all bound to unique ip's). Which server to pick, maybe the $SCM one is
> the right server to contact here.
>
Well, but all of the above doesn't help - using volutil purge ... directly 
simply reports that such a volume does not exist. Therefor I conclude that 
purgevol_rep partially did it's job - but somewhere information about the 
volume remains stored (probably the rvmlog?). I should add, that these 
problems started with version 6.0.7, but obviously there was no problem 
when using 6.0.6.

What has changed around "volutil purge"?

Thanks,
Michael