[Gluster-users] Can I create a distributed dispersed volume with RAM drives?

[Ewen Chan]

Vijay:

I just tried that and the results using this command:

# for i in `seq -w 1 4`; do dd if=/dev/zero of=10Gfile$1 bs=1024k count=10240; done

is about 1/4th the write performance of a distributed volume.

(I'm currently testing with a 64 GB RAM drive on each of my four compute nodes that are then tied together using GlusterFS).

With a distributed volume, and with direct-io-mode=enable, at best, I can execute that same task at around 2 GB/s. (The host system could execute that task at around 2.8 GB/s to /dev/shm.)

With a dispersed volume with 1 brick for redundancy, it drops to around 500 MB/s at best.

Is there no other way to be get the kind of performance that stripping should theorectically provide?

Thank you.

Sincerely,
Ewen

________________________________
From: Vijay Bellur <vbellur at redhat.com>
Sent: August 20, 2019 7:50 PM
To: Ewen Chan <alpha754293 at hotmail.com>
Cc: gluster-users at gluster.org <gluster-users at gluster.org>
Subject: Re: [Gluster-users] Can I create a distributed dispersed volume with RAM drives?



On Tue, Aug 20, 2019 at 4:47 PM Ewen Chan <alpha754293 at hotmail.com<mailto:alpha754293 at hotmail.com>> wrote:
Vijay:

Here is what it says:

[root at node1 brick1]# gluster volume create gv0 stripe 4 node1:/bricks/brick1/gv0 node2:/bricks/brick1/gv0 node3:/bricks/brick1/gv0 node4:/bricks/brick1/gv0
stripe option not supported

Usage:
volume create <NEW-VOLNAME> [stripe <COUNT>] [replica <COUNT> [arbiter <COUNT>]] [disperse [<COUNT>]] [disperse-data <COUNT>] [red
undancy <COUNT>] [transport <tcp|rdma|tcp,rdma>] <NEW-BRICK>... [force]

It says that the stripe option is not supported, but then in the usage right below it, it gives the correct syntax for how to use the gluster volume create command.


Creating stripe volumes has been deprecated in recent releases. You can create a dispersed volume using the following syntax:

# gluster volume create [disperse [<count>]] [redundancy <count>] <brick-list>

HTH,
Vijay


Your help in this matter is greatly appreciated.

Thank you.

Sincerely,
Ewen

________________________________
From: Vijay Bellur <vbellur at redhat.com<mailto:vbellur at redhat.com>>
Sent: August 20, 2019 6:21 PM
To: Ewen Chan <alpha754293 at hotmail.com<mailto:alpha754293 at hotmail.com>>
Cc: gluster-users at gluster.org<mailto:gluster-users at gluster.org> <gluster-users at gluster.org<mailto:gluster-users at gluster.org>>
Subject: Re: [Gluster-users] Can I create a distributed dispersed volume with RAM drives?



On Tue, Aug 20, 2019 at 8:13 AM Ewen Chan <alpha754293 at hotmail.com<mailto:alpha754293 at hotmail.com>> wrote:
To Whom It May Concern:

I am currently using CentOS 7.6.1810 on four compute nodes, where each node has 128 GB of RAM.

I have created a tmpfs mount point using half of the RAM.

I read the CentOS SIG Gluster Quickstart wiki/document (https://wiki.centos.org/SpecialInterestGroup/Storage/gluster-Quickstart). In there, it was creating a replicated volume.

Can I create a distributed dispersed volume with RAM drives?

I was able to create a distributed volume, but I am having difficulties trying to create a distributed dispersed volume.



Can you provide more details about the problems encountered while creating a distributed dispersed volume?

Thanks,
Vijay
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.gluster.org/pipermail/gluster-users/attachments/20190821/c190303c/attachment.html>