Punit Dambiwal
2014-Dec-05 08:47 UTC
[Gluster-users] GlausterFS guest machine I/O is too slow...
Hi, I have the following setup :- 4* node in distributed replicated replica =2 24*brick in each node.... i am using each brick as 256GB SSD... I have 2*10Gb lan with bonding for storage purpose As i am using SSD but still i am getting the very slow I/O In my guest VM..... [root at centos7-2 ~]# dd if=/dev/zero of=/tmp/output bs=1M count=1k;rm -rf /tmp/output 1024+0 records in 1024+0 records out 1073741824 bytes (1.1 GB) copied, 12.9764 s, 82.7 MB/s [root at centos7-2 ~]# If i run the same command on the host it's around 561MB/s... [root at cpu04 10.10.0.6:_vol1]# dd if=/dev/zero of=/rhev/data-center/mnt/glusterSD/10.10.0.6:_vol1/output bs=1M count=1k;rm -rf /rhev/data-center/mnt/glusterSD/10.10.0.6:_vol1/output 1024+0 records in 1024+0 records out 1073741824 bytes (1.1 GB) copied, 1.91395 s, 561 MB/s [root at cpu04 10.10.0.6:_vol1]# Is there any way to optimize it for KVM virtulization....so i will get the best use of the SSD... Thanks, Punit -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://supercolony.gluster.org/pipermail/gluster-users/attachments/20141205/e3adbaa3/attachment.html>
Punit Dambiwal
2014-Dec-06 03:45 UTC
[Gluster-users] GlausterFS guest machine I/O is too slow...
Can any one help me here.... ??? On Fri, Dec 5, 2014 at 4:47 PM, Punit Dambiwal <hypunit at gmail.com> wrote:> Hi, > > I have the following setup :- > > 4* node in distributed replicated > replica =2 > 24*brick in each node.... > i am using each brick as 256GB SSD... > I have 2*10Gb lan with bonding for storage purpose > > As i am using SSD but still i am getting the very slow I/O In my guest > VM..... > > [root at centos7-2 ~]# dd if=/dev/zero of=/tmp/output bs=1M count=1k;rm -rf > /tmp/output > 1024+0 records in > 1024+0 records out > 1073741824 bytes (1.1 GB) copied, 12.9764 s, 82.7 MB/s > [root at centos7-2 ~]# > > If i run the same command on the host it's around 561MB/s... > > [root at cpu04 10.10.0.6:_vol1]# dd if=/dev/zero > of=/rhev/data-center/mnt/glusterSD/10.10.0.6:_vol1/output bs=1M > count=1k;rm -rf /rhev/data-center/mnt/glusterSD/10.10.0.6:_vol1/output > 1024+0 records in > 1024+0 records out > 1073741824 bytes (1.1 GB) copied, 1.91395 s, 561 MB/s > [root at cpu04 10.10.0.6:_vol1]# > > Is there any way to optimize it for KVM virtulization....so i will get the > best use of the SSD... > > Thanks, > Punit >-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://supercolony.gluster.org/pipermail/gluster-users/attachments/20141206/7425a705/attachment.html>