[sheepdog-users] sheepdog with corosync crashed

Vasiliy Tolstov v.tolstov at selfip.ru
Wed Jun 24 08:10:45 CEST 2015


2015-06-23 15:31 GMT+03:00 Hitoshi Mitake <mitake.hitoshi at gmail.com>:
> You can't use ec cluster without enough number of nodes. You need x+y
> nodes for x:y.
>
> Below error is caused by corosync driver. sheepdog cannot handle
> network partition because virtual disks must not be inconsistent
> (other types of storage, e.g. kvs, can handle partition correctly). So
> the driver caused the panic and stopped sheep intentionally.


Why sheepdog exists, but not goes to some cycle to check cluster
members and if it size goes to normal - continue to work?
If i want to modify current behavior to something above does this acceptable?

-- 
Vasiliy Tolstov,
e-mail: v.tolstov at selfip.ru


More information about the sheepdog-users mailing list