Gluster users - Nov 2016 - help

Hi all,

I have a gluster volume with 4 bricks(srv1, srv2, srv3 and srv4). srv2 is
replicate of srv1 and srv3 is replicate of srv3. each of this bricks has
1.7TB data.
I am gonna replace srv2 and srv4
? with two new servers(srvP2 and srvP4). srvP2 and srvP4 are in another
datacenter and as I said each brick has 1.7TB so is there anything that I
should be aware of when I do the replace-brick command? I
 am just worried that may be any problem happened because my data is  a lot.

Bests,
Fariborz?
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://www.gluster.org/pipermail/gluster-users/attachments/20161105/6d715ec3/attachment.html>

On 11/05/2016 05:47 AM, Fariborz Mafakheri wrote:
> Hi all,
>
> I have a gluster volume with 4 bricks(srv1, srv2, srv3 and srv4). srv2 
> is replicate of srv1 and srv3 is replicate of srv3. each of this 
> bricks has 1.7TB data.
> I am gonna replace srv2 and srv4
> ? with two new servers(srvP2 and srvP4). srvP2 and srvP4 are in 
> another datacenter and as I said each brick has 1.7TB so is there 
> anything that I should be aware of when I do the replace-brick command? I
>  am just worried that may be any problem happened because my data is 
>  a lot.
>
The most common problem with that kind of change is networking. People 
forget to open up firewalls, don't have routes from their clients, etc.

When you mirror between datacenters, the clients will be connecting to 
both. This will amplify any latency so if that's bad your writes, 
lookups, etc. will suffer.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://www.gluster.org/pipermail/gluster-users/attachments/20161105/198e781a/attachment.html>