[sheepdog-users] Several difficulties with sheepdog (from 0.4.0-0+tek2b-10 deb package)
nimrodxx at gmx.de
Sun Aug 26 09:55:56 CEST 2012
Am 2012-07-27 12:02, schrieb David Douard:
> Done (thanks for the feedback).
I have rewritten the first part to reflect the zones concept.
Could you please have a look, if its OK, before change it in
== General remarks ==
Sheepdog uses a zone concept for data replication. In the default
configuration, each node in the cluster is one zone and sheepdog
the vdi objects to the number of zones defined by the number of copies
To keep a working cluster, never kill sheeps in more than X-1 zones at
a time, or during recovery is not finished. If you kill X or more zones
at a time, you will have no access to some of the objects. It never
if you kill the complete zone, or only a few sheeps in X different
at a time, there will '''always''' some objects, that are only hold by
these specific sheeps, only the amount of inaccessible objects will
Including up to version 0.4 you will '''loose your data''' in this
With the introduction of plain_store and the rework of farm to use it
core in version 0.5, you wont loose data, but as long as the zones are
restarted, you cant access the objects that are only available on this
zones and your VM get a I/O Error.
More information about the sheepdog-users