[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: [PATCH] RHEL 5.3 #468431 - assure consistency of VG to be removed



The only thing that scares me about this patch is the chance that anaconda will traceback on one of those vgremove or lvremove where otherwise it would have continued on its marry way.

OTOH, if we fail on either one of those we will probably see some other issue further on.  And its probably a good idea to see the traceback sooner than later.  From this point of view it seems OK to me.

Regards

----- "Radek Vykydal" <rvykydal redhat com> wrote:

> Hi,
> 
> this patch is a successor of commit 
> b5a48bfc44a8084b4c751e192c70eb1837e44e19
> which fixed only one part of the bug. This one adds these things:
> 
> - calls vgreduce before removing VG with vgremove
> to put the VG into consistent state (e.g. it removes missing
> PVs from VG )
> - adds verbose to lvm calls for better logging
> - removes silent catching of exceptions
>  (this is only update of the first patch of the bug)
> 
> Note:
> While working on the bug I used dirty hack patch to append to lvmout
> log
> instead of rewriting the file with each command
> (it changed opening mode in execWithRedirect if filename contained
> lvmout).
> I think it is worth considering to do this (lvm bugs can be very hard
> to reproduce so each piece of logged info can be very valuable) in
> some
> cleaner way like adding append kwd param to execWithRedirect or so.
> This would likely require also something to separate outputs of
> commands
> in the file. But it would be for 5.4 I guess.
> 
> Radek
> 
> _______________________________________________
> Anaconda-devel-list mailing list
> Anaconda-devel-list redhat com
> https://www.redhat.com/mailman/listinfo/anaconda-devel-list

-- 
Joel Andres Granados
Red Hat / Brno Czech Republic


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]