Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
On Thu, May 10, 2007 at 02:38:23PM +0100, Mark Watts wrote:
>
> > On Thu, May 10, 2007 at 02:21:53PM +0100, Mark Watts wrote:
> > > > On Thu, May 10, 2007 at 01:05:54PM +0100, Mark Watts wrote:
> > > > > > > It can be one of the following three:
> > > > > > >
> > > > > > > C: write IO is reported as completed, if we know it has
> > > > > > > reached _both_ local and remote DISK.
> > > > > > > * for critical transactional data.
> > > > > >
> > > > > > * for most cases.
> > > > >
> > > > > Most ?
> > > >
> > > > what is your question?
> > >
> > > C: write IO is reported as completed, if we know it has
> > > reached _both_ local and remote DISK.
> > >
> > > * for most cases.
> > >
> > > In what circumstances does this not hold to be true (that data is on the
> > > platters of both disks?)
> >
> > 1. misunderstanding:
> > in that post, read the "*" as "this setting is recommended ..."
>
> No, just cut'n'paste from the drbd.conf file :o)
well, yes, which uses it in this meaning.
> > 2.
> > the statement holds not true when you lose connectivity,
> > obviously: if you cannot communicate with the peer,
> > it cannot write your data.
>
> Ah I see. I half expected you to be referring to some kind of rare
> corner-case.
no. I did not refer to anything.
I meant
protocol C: write IO is reported as completed, if we know it has
reached _both_ local and remote DISK.
this setting is neccessary for critical transactional data.
this setting is recommended for most cases, anyways.
this is the meaning of those "*" in the drbd.conf file.
we should really cleanup our example config and documentation,
and add a few extra words here and there to clarify explicitly
what we mean...
--
: Lars Ellenberg Tel +43-1-8178292-0 :
: LINBIT Information Technologies GmbH Fax +43-1-8178292-82 :
: Vivenotgasse 48, A-1120 Vienna/Europe http://www.linbit.com :
__
please use the "List-Reply" function of your email client.