Re: [Gluster-devel] Whole failure when one glusterfsd brought down.

[Krishna Srinivas]

Hi Henrikas,

"option readdir-force-success on" affects only readdir() syscall.
This change is present in glusterfs-1.3.0-pre2.3

And for your setup, have a look at the wiki doc here:
http://www.gluster.org/docs/index.php/GlusterFS_User_Guide#AFR_Example_in_Clustered_Mode

i.e "AFR in clustered mode"

Let us know if you need any help.

Regards
Krishna

> Hi Krishna,
> Sorry to bother you directly, but it seems what your mailinglist is
> currently down, i writting due to "option readdir-force-success on" You
> wrote what it is "a recent change" i wonder is this change added to
> glusterfs-1.3.0-pre2.3 ?
> And from its naming i guess what it is only for reads, what about writes?
> I want to implement 3 server setup, where files would be mirrored twice,
> so if one server were down system would be fully functional for reading
> and writting files, currently I experience same error as Dale... is it
> possible to achieve such result at all?
>
> IS administratorius
> Henrikas Jureviè
>
>
> UAB "BALTIC DATA CENTER"
> IT paslaugos jûsø verslui
> __________________________
> Þirmûnø 141, LT-09128 Vilnius
> Tel. +370 5 2748294
> Mob.tel. +370 618 81236
> www.bdc.lt
>
>
>