[Gluster-users] Help me please with replication
pooster at gmail.com
Thu May 21 07:28:59 UTC 2009
Basically, you're being affected by the self-healing features of the
replication translator. From the GlusterFS FAQ:
Q: What about deletion self/auto healing?
A: With auto healing or self healing only file creation is healed. If
a brick is missing because of a disk crash re-creation of files is ok
but if it's a temporary network problem synchronizing deletion is
In other words, when that downed node comes back up, the healing
system sees that it has data that its partner lacks. Giving the
benefit of the doubt that it's better to have an extra file laying
around than to suffer the possibility that the file should not have
been deleted, the healer errs on the side of caution and replicates
That said, the replication translator is not really meant to implement
a true RAID1 mirroring scheme, but rather to emulate the good bits of
such a scheme. At least, that's my understanding thereof.
> If both servers up, everything works fine. But if second server down (due
> lost network connection), I run into the problem.
> If I deleted a file from first server and second server will be up, the
> deleted file will be recreated at the first server!
> As you see, this is not RAID1.
> Please help me. Can GlusertFS work as completely RAID1 or no?
SO not teh 1337
More information about the Gluster-users