[Gluster-users] delete brick / format / add empty brick

Pranith Kumar Karampuri pkarampu at redhat.com
Thu Jan 9 03:47:40 UTC 2014


Entry self-heal of big-directories can cause this CPU hike. That is the only thing I can think of.

Pranith
----- Original Message -----
> From: "Lalatendu Mohanty" <lmohanty at redhat.com>
> To: "Alun James" <ajames at tibus.com>, "Ravishankar N" <ravishankar at redhat.com>
> Cc: gluster-users at gluster.org
> Sent: Wednesday, January 8, 2014 8:29:25 PM
> Subject: Re: [Gluster-users] delete brick / format / add empty brick
> 
> On 01/08/2014 08:10 PM, Alun James wrote:
> 
> 
> 
> Is it normal to expect very high server load and clients being unable to
> access the mounts during this process? It means the application running on
> this will need to be offline for hours.
> 
> 
> Nope. I never seen glusterfsd reaching 200% cpu . However it depends on the
> server hardware configuration also. what is the RAM size of your servers?
> Also I will suggest you to check logs and see anything unusual happening.
> 
> 
> 
> 
> 
> 
> From: "Ravishankar N" <ravishankar at redhat.com>
> To: "Alun James" <ajames at tibus.com>
> Cc: gluster-users at gluster.org
> Sent: Wednesday, 8 January, 2014 2:37:05 PM
> Subject: Re: [Gluster-users] delete brick / format / add empty brick
> 
> On 01/08/2014 05:57 PM, Alun James wrote:
> 
> 
> 
> I have this a go.
> 
> gluster volume add-brick myvol replica 2 server2:/brick1
> gluster volume heal myvol full
> 
> It seems to be syncing the files but very slowly. Also the server load on
> server01 has risen to 200+ and the gluster clients are no longer able to
> access the mounts. Is there a way to do this that is not as impactful? Could
> I manually rsync the bricks before adding the second node back in?
> 
> The recommended way to heal is using the command mentioned. The gluster self
> heal daemon takes appropriate file locks before healing. Since clients are
> accessing the volume, I don't think bypassing that and rsyncing the bricks
> is a good idea.
> 
> Regards,
> Ravi
> 
> 
> 
> 
> 
> 
> Alun.
> 
> 
> From: "Ravishankar N" <ravishankar at redhat.com>
> To: "Alun James" <ajames at tibus.com>
> Cc: gluster-users at gluster.org
> Sent: Wednesday, 8 January, 2014 4:00:44 AM
> Subject: Re: [Gluster-users] delete brick / format / add empty brick
> 
> On 01/07/2014 09:40 PM, Alun James wrote:
> 
> 
> 
> Hi folks,
> 
> I had a 2 node (1 brick each) replica, some network meltdown issues seemed to
> cause problems with second node (server02). glusterfsd process reaching
> 200-300% and errors relating to split brain possibilities and self heal
> errors.
> 
> Original volume info:
> 
> Volume Name: myvol
> Type: Replicate
> Status: Started
> Number of Bricks: 2
> Transport-type: tcp
> Bricks:
> Brick1: server01:/brick1
> Brick2: server02:/brick1
> 
> I removed the second brick (that was showing server problems).
> 
> gluster volume remove-brick myvol replica 1 server02:/brick1
> 
> Now the volume status is:
> 
> Volume Name: tsfsvol0
> Type: Distribute
> Status: Started
> Number of Bricks: 1
> Transport-type: tcp
> Bricks:
> Brick1: server01:/brick1
> 
> All is fine and the data on working server is sound.
> 
> The xfs partition for server02:/brick1 has been formatted and therefore the
> data gone. All other gluster config data has remained untouched. Can I
> re-add the second server to the volume with an empty brick and the data will
> auto replicate over from the working server?
> 
> gluster volume add-brick myvol replica 2 server2:/brick1 ??
> 
> 
> Yes this should work fine. You will need to run a `gluster volume heal myvol
> full` to manually trigger the replication.
> 
> 
> 
> 
> 
> 
> 
> ALUN JAMES
> Senior Systems Engineer
> Tibus
> 
> T: +44 (0)28 9033 1122
> E: ajames at tibus.com
> W: www.tibus.com
> 
> Follow us on Twitter @tibus
> 
> Tibus is a trading name of The Internet Business Ltd, a company limited by
> share capital and registered in Northern Ireland, NI31325. It is part of UTV
> Media Plc.
> 
> This email and any attachment may contain confidential information for the
> sole use of the intended recipient. Any review, use, distribution or
> disclosure by others is strictly prohibited. If you are not the intended
> recipient (or authorised to receive for the recipient), please contact the
> sender by reply email and delete all copies of this message.
> 
> 
> _______________________________________________
> Gluster-users mailing list Gluster-users at gluster.org
> http://supercolony.gluster.org/mailman/listinfo/gluster-users
> 
> 
> 
> 
> 
> 
> _______________________________________________
> Gluster-users mailing list Gluster-users at gluster.org
> http://supercolony.gluster.org/mailman/listinfo/gluster-users
> 
> 
> _______________________________________________
> Gluster-users mailing list
> Gluster-users at gluster.org
> http://supercolony.gluster.org/mailman/listinfo/gluster-users



More information about the Gluster-users mailing list