Re: [Gluster-users] Is transport=rdma tested with "stripe"?

[Hatazaki, Takao]

Forgot to mention that I was using CentOS7.3 and GlusterFS 3.10.3 that is the 
latest available.

From: gluster-users-boun...@gluster.org 
[mailto:gluster-users-boun...@gluster.org] On Behalf Of Hatazaki, Takao
Sent: Tuesday, August 15, 2017 2:32 AM
To: gluster-users@gluster.org
Subject: [Gluster-users] Is transport=rdma tested with "stripe"?

Hi,

I have 2 servers with Mellanox InfiniBand FDR hardware/software installed.  A 
volume with "replica 2 transport rdma" works (create on servers, mount and test 
on clients) ok.  A volume with "stripe 2 transport tcp" works ok, too.  A 
volume with "stripe 2 transport rdma" created ok, and mounted ok on a client, 
but writing a file caused "endpoint not connected" error on the client.  Is 
this supposed to work or not tested?

I wanted to check with experts before checking logs and other details.

Regards,

Takao

___
Gluster-users mailing list
Gluster-users@gluster.org
http://lists.gluster.org/mailman/listinfo/gluster-users

[Gluster-users] Is Intel Omni-Path tested/supported?

[Hatazaki, Takao]

Hi,

I am new to GlusterFS.  I have 2 computers with Intel Omni-Path 
hardware/software installed.  RDMA is running ok on those according to 
"ibv_devices" output.  GlusterFS "server" installed and set up ok on those.  
Peer probes on TCP-over-OPA worked ok.  I created a replicated volume with 
transport=tcp, then tested them by mounting the volume on one of servers.  
Things worked perfectly.  Then I changed transport to RDMA.  The volume started 
ok, and the mount command on the same one of servers completed without any 
errors.  However, the servers got hang after a while.  I was wondering if Intel 
Omni-Path hardware was really supported by the latest version...  I was using 
CentOS7.3 and GlustreFS 3.10.3.

Takao

___
Gluster-users mailing list
Gluster-users@gluster.org
http://lists.gluster.org/mailman/listinfo/gluster-users

[Gluster-users] Is transport=rdma tested with "stripe"?

[Hatazaki, Takao]

Hi,

I have 2 servers with Mellanox InfiniBand FDR hardware/software installed.  A 
volume with "replica 2 transport rdma" works (create on servers, mount and test 
on clients) ok.  A volume with "stripe 2 transport tcp" works ok, too.  A 
volume with "stripe 2 transport rdma" created ok, and mounted ok on a client, 
but writing a file caused "endpoint not connected" error on the client.  Is 
this supposed to work or not tested?

I wanted to check with experts before checking logs and other details.

Regards,

Takao

___
Gluster-users mailing list
Gluster-users@gluster.org
http://lists.gluster.org/mailman/listinfo/gluster-users

[Gluster-users] 答复: Mess code when encryption

[Liu, Dan]

HI

I have filed an issue to the bugzilla, please refer to this hyper-link.

https://bugzilla.redhat.com/show_bug.cgi?id=1477014


Regards

发件人: Mohammed Rafi K C [mailto:rkavu...@redhat.com]
发送时间: 2017年7月27日 22:02
收件人: Liu, Dan/刘 丹; gluster-users@gluster.org
主题: Re: [Gluster-users] Mess code when encryption


It looks like a bug to me. Can you file an issue in github.

Regards

Rafi KC


On 07/05/2017 03:30 PM, Liu, Dan wrote:
Hi everyone,

I have a question about using encryption in Gluster FS.

1.Created a file (file size is smaller than 1k) in the volume’s mount point.
2.Read the file, finally got a mess code.
I found that the content I got is from cache and no decryption is operated on 
the file content, so mess code returns.

If I set the following property to off, then everything is OK.

performance.quick-read

performance.write-behind

performance.open-behind



My question is :

   Is above metioned phenomenon right?

   Does I have wrong configurations?

OS : CentOS 7.1
Gluster FS:  3.10.3

Configuration
   According quick start of official website, set features.encryption to on and 
encryption.master-key, then start and mount the volume.


Looking forward to your answers. Thanks.




___

Gluster-users mailing list

Gluster-users@gluster.org

http://lists.gluster.org/mailman/listinfo/gluster-users



___
Gluster-users mailing list
Gluster-users@gluster.org
http://lists.gluster.org/mailman/listinfo/gluster-users

[Gluster-users] [gluster-users] Last call on Gluster Summit CFP - 15 Aug

[Amye Scavarda]

Hi all!

We'll be closing our CFP for Gluster Summit tomorrow at midnight Pacific.
If you've got anything you wanted to submit, go ahead and get to it now!
https://www.gluster.org/events/summit2017/ has all the details.

- amye

-- 
Amye Scavarda | a...@redhat.com | Gluster Community Lead
___
Gluster-users mailing list
Gluster-users@gluster.org
http://lists.gluster.org/mailman/listinfo/gluster-users