[Gluster-users] Possible to use gluster w/ email services + Tuning for fast replication

Tue Nov 2 02:38:31 UTC 2010

On Tuesday 02 November 2010 10:04:27 Rick King wrote:
> Ed, thank you for your response!
> 
> >> Are you examining the second node directly, ie not by mounting it?
> 
> This is an interesting question. I am just examining the 2nd node directly,
> it wasn't obvious to me that the 2nd node needed to mount the data from
> the 1st node. I was just merely expecting the data to be replicate to the
> 2nd node. So my rationale is thinking I should run the following command
> on the 2nd node:
> 
> mount -t glusterfs hostnameA:/test /mnt
> 

You do not need to mount the data from the first node in the second.  As I 
understand GlusterFS  works on the client side. What this means is that you 
must mount the volume in a client machine using either the glusterfs or native 
NFS drivers  and then when you add a file to the mounted volume it will be 
replicated to both nodes. 

Writing directly on one of the server nodes storage (e.g. not through a mount 
point) will also replicate the file eventually due to GlusterFS self-heal 
mechanism but this will take longer to take effect. You can always force the 
replication with the volume rebalance command that is what you are seeing.

I had a similar problem with data not being replicated even when using a 
volume mount point and the problem was that the glusterd daemon was not 
running in one of the nodes.  Make sure the daemon is started:

  /etc/init.d/glusted start

also make sure you configure it to start on system boot (it is not by default). 
You can check the manual for details:

http://www.gluster.com/community/documentation/index.php/Gluster_3.1:_Configuring_glusterd_to_Start_Automatically

> 
> The commands I used to create the volume from the server (hostnameA)
> 
> 1) gluster volume create test replica 2 transport tcp hostnameA:/opt
> hostnameB:/opt
> 
> 2) gluster volume start test
> 
> 3) mount -t glusterfs hostnameA:/test /mnt
> 
Again this third step is not necessary. Make sure the glusterd daemons are 
running in all nodes and that you are accessing the volume through the 
glusterfs client or native NFS client.

> Someone sent a message regarding a tutorial that I haven't read yet, so I
> am going to work through that tutorial, and see if I can answer some of my
> own questions. :)
> 
> Thank you again Ed for the tidbit regarding the latency issue, and your
> comment regarding HTPC applications.
> 
> ~~Rick
> ----- Original Message -----
> From: "Ed W" <lists at wildgooses.com>
> To: "Gluster Users" <gluster-users at gluster.org>
> Sent: Monday, November 1, 2010 2:29:57 PM
> Subject: Re: [Gluster-users] Possible to use gluster w/ email services +
> Tuning for fast replication
> 
> > Right now, I am testing out a 2 node setup, with one server replicating
> > data to another node. One thing I noticed was when I created a file or
> > directory on the server, the new data does not replicate to the other
> > node. The only time data is synced from server to the other node is when
> > I run "gluster volume rebalance test start". Is this normal? I had
> > envisioned gluster would constantly replicate changes from the server to
> > the other nodes, am I off base?
> 
> Are you examining the second node directly, ie not by mounting it?  I
> think the point is that replication only happens when you "observe" the
> second node?
> 
> Glusterfs is targeted for HTPC applications where typically the nodes
> are all connected over high performance interlinks.  It appears that
> performance degrades very quickly as the latency between nodes increases
> and so whether the solution works for you is largely going to be
> determined by the latency between nodes on your network connection.
> 
> I'm not actually sure what some representative numbers should be?  I
> have two machines hooked up using bonded-rr intel gigabit cards
> (crossover to each other) and these ping at around 0.3ms.  However, I
> have one other machine on a gigabit connection, hooked up to a switch
> and that sometimes drops to around 0.15ms...  I believe infiniband will
> drop that latency to some few tens of microseconds?
> 
> So basically every file access on my system would suffer a 0.3ms access
> latency.  This is better than a spining disk with no cache which comes
> in more like 3-10ms, but obviously it's still not brilliant
> 
> Please let us know how you get on?
> 
> Good luck
> 
> Ed W
> _______________________________________________
> Gluster-users mailing list
> Gluster-users at gluster.org
> http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
> DISCLAIMER: This e-mail and any files transmitted with it ('Message') is
> intended only for the use of the recepient (s) named and may contain
> confidential information. Opinions, conclusion and other information in
> this message that do not relate to the official business of King7.
> _______________________________________________
> Gluster-users mailing list
> Gluster-users at gluster.org
> http://gluster.org/cgi-bin/mailman/listinfo/gluster-users

-- 
regards,                                                                                                                                                                                                       
Horacio Sanson