On Thu, Aug 20, 2009 at 11:01 AM, Jeff Blaine<[email protected]> wrote: >> It was determined out of list that the particular error I'm seeing in >> this case is because I was adding a RO volume on the same server but >> different partition as the RW volume. While I know it's terrible >> practice, it did work in previous versions and I was using it for >> testing purposes. Apparently, this is no longer allowed, and gives the >> errors that I'm seeing. > > We experienced the same "surprise!" last week when we tried > to move all of the volumes on one fileserver to another > in order to upgrade the former. > > The destination server already had its own replicas... of > some of the volumes being moved. This had always been allowed > in the past for us.
The risk comes from serving stale data. The VLDB allows just one volume of each type per server. The fileserver doesn't communicate with the VLDB about volumes: it reads the disk. So it has to guess which copy's the "right" one. What happens if it guesses wrong? You report a bug because your releases are "losing your data" that's all still there....you just don't see it. It was only somewhat allowed in the past, anyway; a cross-device link error was already returned in some cases. the difference is now the behavior is consistent. _______________________________________________ OpenAFS-info mailing list [email protected] https://lists.openafs.org/mailman/listinfo/openafs-info
