[sheepdog] [sheepdog/sheepdog] ae04ea: sheep: avoid diskfull caused by recovery process

Hitoshi Mitake mitake.hitoshi at lab.ntt.co.jp
Wed Oct 28 08:15:26 CET 2015


Hi Bastian,

> -----Original Message-----
> From: Bastian Scholz [mailto:nimrodxx at gmx.de]
> Sent: Wednesday, October 28, 2015 4:10 PM
> To: Hitoshi Mitake <mitake.hitoshi at lab.ntt.co.jp>
> Subject: Re: [sheepdog] [sheepdog/sheepdog] ae04ea: sheep: avoid diskfull
> caused by recovery process
> 
> Hi Hitoshi,
> 
> what happend if I dont pass -F to sheep? Does it
> still corrupt the cluster?

If the rest members of the cluster doesn't have enough space for recovery,
disk full will happen without the option. It will be hard to recover.

Thanks,
Hitoshi

> 
> Thanks Bastian
> 
> Am 2015-10-28 06:51, schrieb Hitoshi Mitake:
> > Branch: refs/heads/recovery-diskfull
> >   Home:   https://github.com/sheepdog/sheepdog
> >   Commit: ae04ea86feaa262ab256f0820e59c2c838f6ac15
> >
> >
> https://github.com/sheepdog/sheepdog/commit/ae04ea86feaa262ab256f0820e
> 59c2c838f6ac15
> >   Author: Hitoshi Mitake <mitake.hitoshi at lab.ntt.co.jp>
> >   Date:   2015-10-28 (Wed, 28 Oct 2015)
> >
> >   Changed paths:
> >     M sheep/recovery.c
> >     M sheep/sheep.c
> >     M sheep/sheep_priv.h
> >
> >   Log Message:
> >   -----------
> >   sheep: avoid diskfull caused by recovery process
> >
> > sheep can corrupt its cluster by diskfull with recovery process. For
> > avoiding this problem, this patch adds a new option -F to sheep. If
> > this command is passed to the sheep process, every sheep process of
> > the cluster stops itself if there is a possibility of diskfull during
> > recovery.
> >
> > Fixes https://github.com/sheepdog/sheepdog/issues/59
> >
> > Signed-off-by: Hitoshi Mitake <mitake.hitoshi at lab.ntt.co.jp>
> 




More information about the sheepdog mailing list