Re: [dm-devel] [RFC] How to fix system stall on root volume multipath

[Kiyoshi Ueda <k-ueda ct jp nec com>]

Hi Christophe,

On Mon, 19 Nov 2007 01:17:41 +0100, Christophe Varoqui wrote:
> > > So I propose to start playing with your prioritizers-as-lib idea to see
> > > if it's practical.
> > > 
> > > I prepared the following patch to that effect. It is not complete
> > > (actually segfaults, no useful prioritizer ported) but can start fixing
> > > bugs and go where ever your personnal interest leads.
> >
> > I'm sorry I forgot to post the git-cached part of the changeset, i.e.
> > libprio/ files.
> > 
> > There it goes.
> 
> Upstream git hosts the initial libprio/ commit.
> 
> I ported all useful prioritizers, leaving balance_units in the dark.
> (If someone is actually using it, please say so)
> 
> This time, I cared for a bit of testing and the stuff seems to actually
> work.
> 
> Now I'm interested in feedback from the Netapp fault injection team.
> And any other feedback.

Thank you for the libprio coding.  And I'm sorry for my late reply.

I checked the commit and found a bug during my small testing.
I'll send a patch to fix it by another email.


I would like to note that we lost the priority callout feature
completely to fix the stall problem of root multipath.

Although below is just my humble opinion, I guess it was useful
for people who want to use new prioritizers on official distro's
environments until an update package including the prioritizer
is released.
So if such a case often happens, adding other ways to use external
prioritizers (perhaps dynamic loading library module like LVM2)
would be required in the near future.

Thanks,
Kiyoshi Ueda