Hi,
I see that as a security feature ;-)
You can prevent data loss if k chunks are intact, but you don't want
to work with the least required amount of chunks. In a disaster
scenario you can reduce min_size to k temporarily, but the main goal
should always be to get the OSDs back up.
For example, in a replicated pool with size 3 we set min_size to 2 not
to 1, although that would also work if everything is healthy. But it's
risky since there's also a chance that two corrupt PGs overwrite a
healthy PG.
Regards,
Eugen
Zitat von "Clausen, Jörn" <jclausen@xxxxxxxxx>:
Hi!
While trying to understand erasure coded pools, I would have
expected that "min_size" of a pool is equal to the "K" parameter.
But it turns out, that it is always K+1.
Isn't the description of erasure coding misleading then? In a K+M
setup, I would expect to be good (in the sense of "no service
impact"), even if M OSDs are lost. But in reality, my clients would
already experience an impact when M-1 OSDs are lost. This means, you
should always consider one more spare than you would do in e.g. a
classic RAID setup, right?
Joern
--
Jörn Clausen
Daten- und Rechenzentrum
GEOMAR Helmholtz-Zentrum für Ozeanforschung Kiel
Düsternbrookerweg 20
24105 Kiel
_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com