[sheepdog-users] Testing snapshot-object-reclaim

[Valerio Pachera]

2014-02-27 12:10 GMT+01:00 Valerio Pachera <sirio81 at gmail.com>:
> I'm going to repeat the test with master branch.

The crash doesn't occur on master.

How to reproduce it:
- use snapshot-object-reclaim
- cluster format -c 2:1
- import qcow2
- vdi snapshot
- write a big file till the end of the vdi
- snapshot
- remove the big file
- snapshot
- write agin the big file

At ths point the guest gives I/O error and, on host side, each node
disconnect form the others.
Errors in sheep.log are reported in the previous mail.