[sheepdog] [PATCH v2 0/7] md: add hot-plug and hot-unplug support

MORITA Kazutaka morita.kazutaka at lab.ntt.co.jp
Wed Apr 3 04:07:44 CEST 2013

At Fri, 29 Mar 2013 13:49:37 +0800,
Liu Yuan wrote:
> From: Liu Yuan <tailai.ly at taobao.com>
> v2:
>  - fix object stale purging for md
> This is the final part for MD. With this patch set, We even allow group plug, 
> group unplug and disks/node failure during (un)plugging.
> Also add disk information function for collie.
> Liu Yuan (7):
>   md: add support for simultaneous disk failure on the same node
>   tests/055: test simultaneous multiple disk failures on the same node
>   net: enlarge connect_to_addr() scope
>   md: add hot-plug and hot-unplug support
>   collie: add new commands to manipulate multi-disks
>   sheep: remove duplicate recovery complete notification
>   tsets: add 057 to test md hot-plug and hot-unplug

I ran the followng script:

for i in 0 1 2; do
    ./sheep/sheep /store/$i/,/store/$i/d0,/store/$i/d1 -z $i -p 700$i

sleep 1
./collie/collie cluster format -c 2
./collie/collie vdi create test 200M -P

rm -rf /store/0/d0
rm -rf /store/1/d0
./collie/collie vdi check test

then got the below result.

  using backend farm store
  FATAL: failed to read, Invalid error code

I'm not sure which patch is the root cause yet.



More information about the sheepdog mailing list