Testing oVirt with Gluster 6
Hi, according to https://bugzilla.redhat.com/show_bug.cgi?id=1673058 the network throughput usage increases by a x5 upgrading from Gluster 3.12 to Gluster 5.y. This seems definitively problematic on our hyper-converged environments: I read more than one report where, over 1 Gbps network, sanlock could not be refreshed quickly enough and so killed the engine VM if not rebooted the host due to the watchdog. This is probably due to the fact that, according to that bug, the network got completely saturated. Faster networks seems still fine. https://bugzilla.redhat.com/show_bug.cgi?id=1673058 is probably going to be solved for Gluster 5.6. On the other side we have https://bugzilla.redhat.com/1693998 <https://bugzilla.redhat.com/show_bug.cgi?id=1693998> about rebasing over Gluster 6 which seems not affected by that issue. Do you have concrete plans about testing oVirt with Gluster 6? thanks, Simone
Hi Simone, Kaustav has upgraded from gluster-3.12 to gluster-6 and done some tests like VM creation and some pg bench testing. https://github.com/ezio-auditore/gluster-pg-bench-test-results But yeah as you said will have to plan for end to end testing with fresh installation also. On Fri, Apr 5, 2019 at 3:31 PM Simone Tiraboschi <stirabos@redhat.com> wrote:
Hi, according to https://bugzilla.redhat.com/show_bug.cgi?id=1673058 the network throughput usage increases by a x5 upgrading from Gluster 3.12 to Gluster 5.y.
This seems definitively problematic on our hyper-converged environments: I read more than one report where, over 1 Gbps network, sanlock could not be refreshed quickly enough and so killed the engine VM if not rebooted the host due to the watchdog. This is probably due to the fact that, according to that bug, the network got completely saturated. Faster networks seems still fine. https://bugzilla.redhat.com/show_bug.cgi?id=1673058 is probably going to be solved for Gluster 5.6.
On the other side we have https://bugzilla.redhat.com/1693998 <https://bugzilla.redhat.com/show_bug.cgi?id=1693998> about rebasing over Gluster 6 which seems not affected by that issue. Do you have concrete plans about testing oVirt with Gluster 6?
thanks, Simone
-- Thanks, Gobinda
Il giorno ven 5 apr 2019 alle ore 12:56 Gobinda Das <godas@redhat.com> ha scritto:
Hi Simone, Kaustav has upgraded from gluster-3.12 to gluster-6 and done some tests like VM creation and some pg bench testing. https://github.com/ezio-auditore/gluster-pg-bench-test-results But yeah as you said will have to plan for end to end testing with fresh installation also.
Gobinda, Sahina, Kaustav, any update on this testing? Are we ok to go with pushing the switch to Gluster 6 to master? One month passed and it's time to go on with this.
On Fri, Apr 5, 2019 at 3:31 PM Simone Tiraboschi <stirabos@redhat.com> wrote:
Hi, according to https://bugzilla.redhat.com/show_bug.cgi?id=1673058 the network throughput usage increases by a x5 upgrading from Gluster 3.12 to Gluster 5.y.
This seems definitively problematic on our hyper-converged environments: I read more than one report where, over 1 Gbps network, sanlock could not be refreshed quickly enough and so killed the engine VM if not rebooted the host due to the watchdog. This is probably due to the fact that, according to that bug, the network got completely saturated. Faster networks seems still fine. https://bugzilla.redhat.com/show_bug.cgi?id=1673058 is probably going to be solved for Gluster 5.6.
On the other side we have https://bugzilla.redhat.com/1693998 <https://bugzilla.redhat.com/show_bug.cgi?id=1693998> about rebasing over Gluster 6 which seems not affected by that issue. Do you have concrete plans about testing oVirt with Gluster 6?
thanks, Simone
--
Thanks, Gobinda _______________________________________________ Devel mailing list -- devel@ovirt.org To unsubscribe send an email to devel-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/devel@ovirt.org/message/RFNOOVJAYJNQ7Q...
-- Sandro Bonazzola MANAGER, SOFTWARE ENGINEERING, EMEA R&D RHV Red Hat EMEA <https://www.redhat.com/> sbonazzo@redhat.com <https://red.ht/sig> <https://redhat.com/summit>
Hi, We have tested gluster 6 by running pg bench on a vm. Things are working fine. Here are the results https://github.com/ezio-auditore/gluster-pg-bench-test-results There is a small blog regarding this which is yet to be published. You can check the preview. https://ezio-auditore.github.io/devblog/gluster-3-12-vs-6-a-performance-orie... On Thu, May 9, 2019 at 8:03 PM Sandro Bonazzola <sbonazzo@redhat.com> wrote:
Il giorno ven 5 apr 2019 alle ore 12:56 Gobinda Das <godas@redhat.com> ha scritto:
Hi Simone, Kaustav has upgraded from gluster-3.12 to gluster-6 and done some tests like VM creation and some pg bench testing. https://github.com/ezio-auditore/gluster-pg-bench-test-results But yeah as you said will have to plan for end to end testing with fresh installation also.
Gobinda, Sahina, Kaustav, any update on this testing? Are we ok to go with pushing the switch to Gluster 6 to master? One month passed and it's time to go on with this.
On Fri, Apr 5, 2019 at 3:31 PM Simone Tiraboschi <stirabos@redhat.com> wrote:
Hi, according to https://bugzilla.redhat.com/show_bug.cgi?id=1673058 the network throughput usage increases by a x5 upgrading from Gluster 3.12 to Gluster 5.y.
This seems definitively problematic on our hyper-converged environments: I read more than one report where, over 1 Gbps network, sanlock could not be refreshed quickly enough and so killed the engine VM if not rebooted the host due to the watchdog. This is probably due to the fact that, according to that bug, the network got completely saturated. Faster networks seems still fine. https://bugzilla.redhat.com/show_bug.cgi?id=1673058 is probably going to be solved for Gluster 5.6.
On the other side we have https://bugzilla.redhat.com/1693998 <https://bugzilla.redhat.com/show_bug.cgi?id=1693998> about rebasing over Gluster 6 which seems not affected by that issue. Do you have concrete plans about testing oVirt with Gluster 6?
thanks, Simone
--
Thanks, Gobinda _______________________________________________ Devel mailing list -- devel@ovirt.org To unsubscribe send an email to devel-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/devel@ovirt.org/message/RFNOOVJAYJNQ7Q...
--
Sandro Bonazzola
MANAGER, SOFTWARE ENGINEERING, EMEA R&D RHV
Red Hat EMEA <https://www.redhat.com/>
sbonazzo@redhat.com <https://red.ht/sig> <https://redhat.com/summit>
-- Thanks, Kaustav Majumder
Il giorno lun 13 mag 2019 alle ore 08:06 Kaustav Majumder < kmajumde@redhat.com> ha scritto:
Hi, We have tested gluster 6 by running pg bench on a vm. Things are working fine. Here are the results https://github.com/ezio-auditore/gluster-pg-bench-test-results
There is a small blog regarding this which is yet to be published. You can check the preview.
https://ezio-auditore.github.io/devblog/gluster-3-12-vs-6-a-performance-orie...
Thanks, I'll push patches for switching to Gluster 6 on master and if no regressions will be found I'll push to 4.3 too.
On Thu, May 9, 2019 at 8:03 PM Sandro Bonazzola <sbonazzo@redhat.com> wrote:
Il giorno ven 5 apr 2019 alle ore 12:56 Gobinda Das <godas@redhat.com> ha scritto:
Hi Simone, Kaustav has upgraded from gluster-3.12 to gluster-6 and done some tests like VM creation and some pg bench testing. https://github.com/ezio-auditore/gluster-pg-bench-test-results But yeah as you said will have to plan for end to end testing with fresh installation also.
Gobinda, Sahina, Kaustav, any update on this testing? Are we ok to go with pushing the switch to Gluster 6 to master? One month passed and it's time to go on with this.
On Fri, Apr 5, 2019 at 3:31 PM Simone Tiraboschi <stirabos@redhat.com> wrote:
Hi, according to https://bugzilla.redhat.com/show_bug.cgi?id=1673058 the network throughput usage increases by a x5 upgrading from Gluster 3.12 to Gluster 5.y.
This seems definitively problematic on our hyper-converged environments: I read more than one report where, over 1 Gbps network, sanlock could not be refreshed quickly enough and so killed the engine VM if not rebooted the host due to the watchdog. This is probably due to the fact that, according to that bug, the network got completely saturated. Faster networks seems still fine. https://bugzilla.redhat.com/show_bug.cgi?id=1673058 is probably going to be solved for Gluster 5.6.
On the other side we have https://bugzilla.redhat.com/1693998 <https://bugzilla.redhat.com/show_bug.cgi?id=1693998> about rebasing over Gluster 6 which seems not affected by that issue. Do you have concrete plans about testing oVirt with Gluster 6?
thanks, Simone
--
Thanks, Gobinda _______________________________________________ Devel mailing list -- devel@ovirt.org To unsubscribe send an email to devel-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/devel@ovirt.org/message/RFNOOVJAYJNQ7Q...
--
Sandro Bonazzola
MANAGER, SOFTWARE ENGINEERING, EMEA R&D RHV
Red Hat EMEA <https://www.redhat.com/>
sbonazzo@redhat.com <https://red.ht/sig> <https://redhat.com/summit>
--
Thanks,
Kaustav Majumder
-- Sandro Bonazzola MANAGER, SOFTWARE ENGINEERING, EMEA R&D RHV Red Hat EMEA <https://www.redhat.com/> sbonazzo@redhat.com <https://red.ht/sig> <https://redhat.com/summit>
Hi all,
On Apr 5, 2019, at 12:00 PM, Simone Tiraboschi <stirabos@redhat.com> wrote:
Hi, according to https://bugzilla.redhat.com/show_bug.cgi?id=1673058 <https://bugzilla.redhat.com/show_bug.cgi?id=1673058> the network throughput usage increases by a x5 upgrading from Gluster 3.12 to Gluster 5.y.
This seems definitively problematic on our hyper-converged environments: I read more than one report where, over 1 Gbps network, sanlock could not be refreshed quickly enough and so killed the engine VM if not rebooted the host due to the watchdog. This is probably due to the fact that, according to that bug, the network got completely saturated. Faster networks seems still fine. https://bugzilla.redhat.com/show_bug.cgi?id=1673058 <https://bugzilla.redhat.com/show_bug.cgi?id=1673058> is probably going to be solved for Gluster 5.6.
On the other side we have https://bugzilla.redhat.com/1693998 <https://bugzilla.redhat.com/show_bug.cgi?id=1693998> about rebasing over Gluster 6 which seems not affected by that issue. Do you have concrete plans about testing oVirt with Gluster 6?
based on bug above, any plan to switch to Gluster-6 with oVirt 4.3 ? Simon
thanks, Simone
_______________________________________________ Devel mailing list -- devel@ovirt.org To unsubscribe send an email to devel-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/devel@ovirt.org/message/IVTYCFKCQ5P2WZ...
Il giorno lun 8 apr 2019 alle ore 19:48 Simon Coter <simon.coter@oracle.com> ha scritto:
Hi all,
On Apr 5, 2019, at 12:00 PM, Simone Tiraboschi <stirabos@redhat.com> wrote:
Hi, according to https://bugzilla.redhat.com/show_bug.cgi?id=1673058 the network throughput usage increases by a x5 upgrading from Gluster 3.12 to Gluster 5.y.
This seems definitively problematic on our hyper-converged environments: I read more than one report where, over 1 Gbps network, sanlock could not be refreshed quickly enough and so killed the engine VM if not rebooted the host due to the watchdog. This is probably due to the fact that, according to that bug, the network got completely saturated. Faster networks seems still fine. https://bugzilla.redhat.com/show_bug.cgi?id=1673058 is probably going to be solved for Gluster 5.6.
On the other side we have https://bugzilla.redhat.com/1693998 <https://bugzilla.redhat.com/show_bug.cgi?id=1693998> about rebasing over Gluster 6 which seems not affected by that issue. Do you have concrete plans about testing oVirt with Gluster 6?
based on bug above, any plan to switch to Gluster-6 with oVirt 4.3 ?
We are tentatively targeting rebase on Gluster 6 for 4.3.5: https://bugzilla.redhat.com/show_bug.cgi?id=1693998 Plan can change while we test it.
Simon
thanks, Simone
_______________________________________________ Devel mailing list -- devel@ovirt.org To unsubscribe send an email to devel-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/devel@ovirt.org/message/IVTYCFKCQ5P2WZ...
_______________________________________________ Devel mailing list -- devel@ovirt.org To unsubscribe send an email to devel-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/devel@ovirt.org/message/2RM2XMW5BV2C7J...
-- SANDRO BONAZZOLA MANAGER, SOFTWARE ENGINEERING, EMEA R&D RHV Red Hat EMEA <https://www.redhat.com/> sbonazzo@redhat.com <https://red.ht/sig>
Thanks Sandro for your feedback. Simon
On Apr 9, 2019, at 4:36 PM, Sandro Bonazzola <sbonazzo@redhat.com> wrote:
Il giorno lun 8 apr 2019 alle ore 19:48 Simon Coter <simon.coter@oracle.com <mailto:simon.coter@oracle.com>> ha scritto: Hi all,
On Apr 5, 2019, at 12:00 PM, Simone Tiraboschi <stirabos@redhat.com <mailto:stirabos@redhat.com>> wrote:
Hi, according to https://bugzilla.redhat.com/show_bug.cgi?id=1673058 <https://bugzilla.redhat.com/show_bug.cgi?id=1673058> the network throughput usage increases by a x5 upgrading from Gluster 3.12 to Gluster 5.y.
This seems definitively problematic on our hyper-converged environments: I read more than one report where, over 1 Gbps network, sanlock could not be refreshed quickly enough and so killed the engine VM if not rebooted the host due to the watchdog. This is probably due to the fact that, according to that bug, the network got completely saturated. Faster networks seems still fine. https://bugzilla.redhat.com/show_bug.cgi?id=1673058 <https://bugzilla.redhat.com/show_bug.cgi?id=1673058> is probably going to be solved for Gluster 5.6.
On the other side we have https://bugzilla.redhat.com/1693998 <https://bugzilla.redhat.com/show_bug.cgi?id=1693998> about rebasing over Gluster 6 which seems not affected by that issue. Do you have concrete plans about testing oVirt with Gluster 6?
based on bug above, any plan to switch to Gluster-6 with oVirt 4.3 ?
We are tentatively targeting rebase on Gluster 6 for 4.3.5: https://bugzilla.redhat.com/show_bug.cgi?id=1693998 <https://bugzilla.redhat.com/show_bug.cgi?id=1693998> Plan can change while we test it.
Simon
thanks, Simone
_______________________________________________ Devel mailing list -- devel@ovirt.org <mailto:devel@ovirt.org> To unsubscribe send an email to devel-leave@ovirt.org <mailto:devel-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/devel@ovirt.org/message/IVTYCFKCQ5P2WZ... <https://lists.ovirt.org/archives/list/devel@ovirt.org/message/IVTYCFKCQ5P2WZ43CM245PTX5T54PLQX/>
_______________________________________________ Devel mailing list -- devel@ovirt.org <mailto:devel@ovirt.org> To unsubscribe send an email to devel-leave@ovirt.org <mailto:devel-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/devel@ovirt.org/message/2RM2XMW5BV2C7J... <https://lists.ovirt.org/archives/list/devel@ovirt.org/message/2RM2XMW5BV2C7J54TRR6W6YUCVPCBPPM/>
-- SANDRO BONAZZOLA MANAGER, SOFTWARE ENGINEERING, EMEA R&D RHV Red Hat EMEA <https://www.redhat.com/> sbonazzo@redhat.com <mailto:sbonazzo@redhat.com>
<https://red.ht/sig> _______________________________________________ Devel mailing list -- devel@ovirt.org <mailto:devel@ovirt.org> To unsubscribe send an email to devel-leave@ovirt.org <mailto:devel-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/devel@ovirt.org/message/TENVEABIF7ERTZ... <https://lists.ovirt.org/archives/list/devel@ovirt.org/message/TENVEABIF7ERTZB243GJXYZAXGNZ7CQU/>
participants (5)
-
Gobinda Das -
Kaustav Majumder -
Sandro Bonazzola -
Simon Coter -
Simone Tiraboschi