[Gluster-devel] Server-side AFR + Failover and mixed server/client-side AFR

Krishna Srinivas krishna at zresearch.com
Tue May 6 15:40:53 UTC 2008


On Tue, May 6, 2008 at 9:09 PM, Krishna Srinivas <krishna at zresearch.com> wrote:
> On Tue, May 6, 2008 at 9:04 PM,  <gordan at bobich.net> wrote:
>  > On Tue, 6 May 2008, Krishna Srinivas wrote:
>  >
>  >
>  > >
>  > > > If server-side AFR breaks with fail-over, is it possible to use
>  > server-side
>  > > >
>  > > > AFR on some nodes and client-side AFR on others, with the client side
>  > AFR
>  > > > nodes talking directly to the same storage bricks that the server side
>  > AFR
>  > > > is unifying for server-side AFR nodes?
>  > > >
>  > > >  Or would this break things?
>  > > >
>  > >
>  > >
>  > > It will not break things as long as the afr's subvol listing order are
>  > same.
>  > >
>  >
>  >  Is this an issue in server-side-only AFR? I have two servers which are also
>  > clients of themselves, and they both list their local subvolume first and
>  > remote subvolume second. Is this a problem? What are the possible
>  > consequences of this?
>
>  It will be a problem. The "first" subvol is always the "lock" server.
>  Consider a case where you are creating a file simultaneously
>  on two clients, only one of them should succeed. If AFR's
>  subvols order are not same, chances are that both client
>  returns success for file creation with same name.

Hence you have "option read-subvlume" to speeden the
read() calls so that it can be done from local subvol.

>
>  Krishna
>
>
>
>
>  >
>  >
>  >
>  >  Gordan
>  >
>  >
>  >  _______________________________________________
>  >  Gluster-devel mailing list
>  >  Gluster-devel at nongnu.org
>  >  http://lists.nongnu.org/mailman/listinfo/gluster-devel
>  >
>





More information about the Gluster-devel mailing list