Hi, Currently, we have three UCS C220 M4, dual Xeon CPU (48 cores), 32GB of RAM, 8x900GB spindles, with Intel X520 dual 10G ports. We are planning to migrate more VMs and increase the number of servers in the cluster as soon as we figure what's going on with the NFS mount. -- Respectfully Mahdi A. Mahdi> From: gandalf.corvotempesta at gmail.com > Date: Wed, 3 Aug 2016 20:25:56 +0200 > Subject: Re: [Gluster-users] Gluster 3.7.13 NFS Crash > To: mahdi.adnan at outlook.com > CC: kdhananj at redhat.com; gluster-users at gluster.org > > 2016-08-03 17:02 GMT+02:00 Mahdi Adnan <mahdi.adnan at outlook.com>: > > the problem is, the current setup is used in a production environment, and > > switching the mount point of +50 VMs from native nfs to nfs-ganesha is not > > going to be smooth and without downtime, so i really appreciate your > > thoughts on this matter. > > A little bit OT: > > 50+ VMs? Could you please share your hardware and network infrastructure? > We are thinking about a gluster cluster for hosting about 80-100 VMs and we are > looking for some production clusters to use as reference.-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://www.gluster.org/pipermail/gluster-users/attachments/20160803/13a3f530/attachment.html>
I had stability problems with centos 7.2 and gluster 3.7.11. Nodes were crashing without any clue. I cannot find a solution and switched to centos 6.8. All problems gone with 6.8. Maybe you can test with centos 6.8? On Wed, Aug 3, 2016 at 10:40 PM, Mahdi Adnan <mahdi.adnan at outlook.com> wrote:> Hi, > > Currently, we have three UCS C220 M4, dual Xeon CPU (48 cores), 32GB of RAM, > 8x900GB spindles, with Intel X520 dual 10G ports. We are planning to migrate > more VMs and increase the number of servers in the cluster as soon as we > figure what's going on with the NFS mount. > > > -- > > Respectfully > Mahdi A. Mahdi > >> From: gandalf.corvotempesta at gmail.com >> Date: Wed, 3 Aug 2016 20:25:56 +0200 >> Subject: Re: [Gluster-users] Gluster 3.7.13 NFS Crash >> To: mahdi.adnan at outlook.com >> CC: kdhananj at redhat.com; gluster-users at gluster.org >> >> 2016-08-03 17:02 GMT+02:00 Mahdi Adnan <mahdi.adnan at outlook.com>: >> > the problem is, the current setup is used in a production environment, >> > and >> > switching the mount point of +50 VMs from native nfs to nfs-ganesha is >> > not >> > going to be smooth and without downtime, so i really appreciate your >> > thoughts on this matter. >> >> A little bit OT: >> >> 50+ VMs? Could you please share your hardware and network infrastructure? >> We are thinking about a gluster cluster for hosting about 80-100 VMs and >> we are >> looking for some production clusters to use as reference. > > _______________________________________________ > Gluster-users mailing list > Gluster-users at gluster.org > http://www.gluster.org/mailman/listinfo/gluster-users
2016-08-03 21:40 GMT+02:00 Mahdi Adnan <mahdi.adnan at outlook.com>:> Hi, > > Currently, we have three UCS C220 M4, dual Xeon CPU (48 cores), 32GB of RAM, > 8x900GB spindles, with Intel X520 dual 10G ports. We are planning to migrate > more VMs and increase the number of servers in the cluster as soon as we > figure what's going on with the NFS mount.Only 3 servers? How many IOPS are you getting and how much bandwidth when reading/writing? 900GB SAS 15k?