[DRBD-user] DRBD 0.7.23 and MD corruption

Ross S. W. Walker rwalker at medallion.com
Wed May 23 19:04:05 CEST 2007

Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.

> -----Original Message-----
> From: Lars Ellenberg [mailto:lars.ellenberg at linbit.com] 
> Sent: Wednesday, May 23, 2007 12:27 PM
> To: Ross S. W. Walker
> Cc: drbd-user at lists.linbit.com
> Subject: Re: [DRBD-user] DRBD 0.7.23 and MD corruption
> On Wed, May 23, 2007 at 12:16:25PM -0400, Ross S. W. Walker wrote:
> > Lars, et al,
> > 
> > I was wondering if a block storage device, which could 
> exist on an MD,
> > were to be fully utilized before it was incorporated into a drbd
> > replication scheme would it also suffer from corruption.
> > 
> > Why I ask, is that I have a bunch of volumes used as iSCSI targets
> > that I am planning on replicating in the future, but are in 
> production
> > today and wanted to know if it is safe to create replicas of these,
> > using a separate LV for meta-data?
> if you have the drbd meta data on a separate lv,
> DRBD is completely "transparent",
> comparable to a dm-linear mapping over the full device.

That is what I figured.

> so it "should just work",
> and it does so for us, all the time.
> but, you know, murphy and "shit happens" ...
> and the issue with the "READA failing without error on MD raid5",
> mentioned earlier in this thread is a real one.
> dm-crypt has suffered from it, and so would drbd, probably,
> when stacked directly on a md raid5 array.

Yes, there seems to always be something creeping up from MD/DM...

I remember the performance issue with meta-data on raid1 and barriers
not being passed down.

> we will double check and realease a workaround.
> note that typical usage is unaffected, though.
> [md is a pv, and drbd resides on an lv, so on dm-linear]

I think the safest thing is to use hardware raid for important data.

Another thing, is there any plans to implement a many-to-many
replication topology with DRBD say utilizing multicasting?


This e-mail, and any attachments thereto, is intended only for use by
the addressee(s) named herein and may contain legally privileged
and/or confidential information. If you are not the intended recipient
of this e-mail, you are hereby notified that any dissemination,
distribution or copying of this e-mail, and any attachments thereto,
is strictly prohibited. If you have received this e-mail in error,
please immediately notify the sender and permanently delete the
original and any copy or printout thereof.

More information about the drbd-user mailing list