RAID L1/L5/L10 + NFS Loopback Benchmarks
Hi ! I run a series of benchmarks with IOZONE testing performance of different RAID configurations on HP ProLiant DL380 + impact of NFS loopback. NFS loopback is necessary even for very small systems (e.g. just 2 nodes), since oVirt local storage setup is limited to single (only one) node per data center and this is quite useless from the point of view of functionality. Full article: http://www.vacuum-tube.eu/wp/?p=537 Short YouTube slide show: https://www.youtube.com/watch?v=Xw_VN9G8jSI Related video: Moving away from local storage setup to NFS/GlusterFS https://www.youtube.com/watch?v=2ebhZ5U6x_Y
On Sun, Nov 18, 2018 at 12:58 PM Andrei Verovski <andreil1@starlett.lv> wrote:
Hi !
I run a series of benchmarks with IOZONE testing performance of different RAID configurations on HP ProLiant DL380 + impact of NFS loopback.
NFS loopback is necessary even for very small systems (e.g. just 2
AFAIK NFS loopback is still considered a bad practice and it's still strongly recommended against since it could lead to deadlock and data loss. See https://access.redhat.com/solutions/22231 I'd strongly recommend a single node hyper-converged gluster over that.
nodes), since oVirt local storage setup is limited to single (only one) node per data center and this is quite useless from the point of view of functionality.
Full article: http://www.vacuum-tube.eu/wp/?p=537
Short YouTube slide show: https://www.youtube.com/watch?v=Xw_VN9G8jSI
Related video: Moving away from local storage setup to NFS/GlusterFS https://www.youtube.com/watch?v=2ebhZ5U6x_Y _______________________________________________ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-leave@ovirt.org Privacy Statement: https://www.ovirt.org/site/privacy-policy/ oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/OCXJAMVDUVLZR5...
Hi ! Deadlock of NFS loopback happens with 3.10 stock kernel from CentOS 7.6 only or also with 4.x mainline ? I use 4.x mainline on my nodes.
On 19 Nov 2018, at 12:59, Simone Tiraboschi <stirabos@redhat.com> wrote:
On Sun, Nov 18, 2018 at 12:58 PM Andrei Verovski <andreil1@starlett.lv <mailto:andreil1@starlett.lv>> wrote: Hi !
I run a series of benchmarks with IOZONE testing performance of different RAID configurations on HP ProLiant DL380 + impact of NFS loopback.
NFS loopback is necessary even for very small systems (e.g. just 2
AFAIK NFS loopback is still considered a bad practice and it's still strongly recommended against since it could lead to deadlock and data loss. See https://access.redhat.com/solutions/22231 <https://access.redhat.com/solutions/22231> I'd strongly recommend a single node hyper-converged gluster over that.
nodes), since oVirt local storage setup is limited to single (only one) node per data center and this is quite useless from the point of view of functionality.
Full article: http://www.vacuum-tube.eu/wp/?p=537 <http://www.vacuum-tube.eu/wp/?p=537>
Short YouTube slide show: https://www.youtube.com/watch?v=Xw_VN9G8jSI <https://www.youtube.com/watch?v=Xw_VN9G8jSI>
Related video: Moving away from local storage setup to NFS/GlusterFS https://www.youtube.com/watch?v=2ebhZ5U6x_Y <https://www.youtube.com/watch?v=2ebhZ5U6x_Y> _______________________________________________ Users mailing list -- users@ovirt.org <mailto:users@ovirt.org> To unsubscribe send an email to users-leave@ovirt.org <mailto:users-leave@ovirt.org> Privacy Statement: https://www.ovirt.org/site/privacy-policy/ <https://www.ovirt.org/site/privacy-policy/> oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ <https://www.ovirt.org/community/about/community-guidelines/> List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/OCXJAMVDUVLZR5... <https://lists.ovirt.org/archives/list/users@ovirt.org/message/OCXJAMVDUVLZR5YUR26GD3CTOM4LJCJZ/>
On Wed, Nov 21, 2018 at 10:08 AM Andrei Verovski <andreil1@starlett.lv> wrote:
Hi !
Deadlock of NFS loopback happens with 3.10 stock kernel from CentOS 7.6 only or also with 4.x mainline ?
I use 4.x mainline on my nodes.
AFAIK is still an open issue.
On 19 Nov 2018, at 12:59, Simone Tiraboschi <stirabos@redhat.com> wrote:
On Sun, Nov 18, 2018 at 12:58 PM Andrei Verovski <andreil1@starlett.lv> wrote:
Hi !
I run a series of benchmarks with IOZONE testing performance of different RAID configurations on HP ProLiant DL380 + impact of NFS loopback.
NFS loopback is necessary even for very small systems (e.g. just 2
AFAIK NFS loopback is still considered a bad practice and it's still strongly recommended against since it could lead to deadlock and data loss. See https://access.redhat.com/solutions/22231 I'd strongly recommend a single node hyper-converged gluster over that.
nodes), since oVirt local storage setup is limited to single (only one) node per data center and this is quite useless from the point of view of functionality.
Full article: http://www.vacuum-tube.eu/wp/?p=537
Short YouTube slide show: https://www.youtube.com/watch?v=Xw_VN9G8jSI
Related video: Moving away from local storage setup to NFS/GlusterFS https://www.youtube.com/watch?v=2ebhZ5U6x_Y _______________________________________________ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-leave@ovirt.org Privacy Statement: https://www.ovirt.org/site/privacy-policy/ oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/OCXJAMVDUVLZR5...
Simone Tiraboschi <stirabos@redhat.com> writes:
On Wed, Nov 21, 2018 at 10:08 AM Andrei Verovski <andreil1@starlett.lv> wrote:
Hi !
Deadlock of NFS loopback happens with 3.10 stock kernel from CentOS 7.6 only or also with 4.x mainline ?
I use 4.x mainline on my nodes.
AFAIK is still an open issue.
FWIW, I've been running with an NFS loopback and have never, in 3+ years, had an issue. On the other hand, I am very over-provisioned and under-subscribed on RAM, so it probably never really gets into a situation where it has to swap. -derek -- Derek Atkins 617-623-3745 derek@ihtfp.com www.ihtfp.com Computer and Internet Security Consultant
participants (3)
-
Andrei Verovski -
Derek Atkins -
Simone Tiraboschi