4:16 p.m.
Hello,
I have a 3 node ovirt 4.3 cluster that I've setup and using gluster (Hyperconverged
setup)
I need to increase the amount of storage and compute so I added a 4th host
(server4.example.com) if it is possible to expand the amount of bricks (storage) in the
"data" volume?
I did some research and from that research the following caught my eye: old post
"https://medium.com/@tumballi/scale-your-gluster-cluster-1-node-at-a-time-62dd6614194e"
So the question is, if taking that approach feasible? , is it even possible an oVirt point
of view?
---------------------------
My gluster volume:
---------------------------
Volume Name: data
Type: Replicate
Volume ID: 003ffea0-b441-43cb-a38f-ccdf6ffb77f8
Status: Started
Snapshot Count: 0
Number of Bricks: 1 x (2 + 1) = 3
Transport-type: tcp
Bricks:
Brick1: server1.example1.com:/gluster_bricks/data/data
Brick2: server2.example.com:/gluster_bricks/data/data
Brick3: server3.example.com:/gluster_bricks/data/data (arbiter)
Options Reconfigured:
cluster.granular-entry-heal: enable
performance.strict-o-direct: on
network.ping-timeout: 30
storage.owner-gid: 36
storage.owner-uid: 36
cluster.choose-local: off
user.cifs: off
features.shard: on
cluster.shd-wait-qlength: 10000
cluster.shd-max-threads: 8
cluster.locking-scheme: granular
cluster.data-self-heal-algorithm: full
cluster.server-quorum-type: server
cluster.quorum-type: auto
cluster.eager-lock: enable
network.remote-dio: enable
performance.low-prio-threads: 32
performance.io-cache: off
performance.read-ahead: off
performance.quick-read: off
transport.address-family: inet
nfs.disable: on
performance.client-io-threads: off
Thanks.
9:52 a.m.
On Thu, Feb 21, 2019 at 8:47 PM <adrianquintero(a)gmail.com> wrote:
Hello,
I have a 3 node ovirt 4.3 cluster that I've setup and using gluster (Hyperconverged
setup)
I need to increase the amount of storage and compute so I added a 4th host
(server4.example.com) if it is possible to expand the amount of bricks (storage) in the
"data" volume?
I did some research and from that research the following caught my eye: old post
"https://medium.com/@tumballi/scale-your-gluster-cluster-1-node-at-a-time-62dd6614194e"
So the question is, if taking that approach feasible? , is it even possible an oVirt
point of view?
Expanding by 1 node is only possible if you have sufficient space on
the existing nodes in order to create a replica 2 + arbiter volume.
The post talks of how you can create your volumes in a way that you
can move the bricks around so that the each of the bricks in a replica
set reside on a separate server. We do not have an automatic way to
provision and rebalance the bricks yet in order to expand by 1 node.
>
>
> ---------------------------
> My gluster volume:
> ---------------------------
> Volume Name: data
> Type: Replicate
> Volume ID: 003ffea0-b441-43cb-a38f-ccdf6ffb77f8
> Status: Started
> Snapshot Count: 0
> Number of Bricks: 1 x (2 + 1) = 3
> Transport-type: tcp
> Bricks:
> Brick1: server1.example1.com:/gluster_bricks/data/data
> Brick2: server2.example.com:/gluster_bricks/data/data
> Brick3: server3.example.com:/gluster_bricks/data/data (arbiter)
> Options Reconfigured:
> cluster.granular-entry-heal: enable
> performance.strict-o-direct: on
> network.ping-timeout: 30
> storage.owner-gid: 36
> storage.owner-uid: 36
> cluster.choose-local: off
> user.cifs: off
> features.shard: on
> cluster.shd-wait-qlength: 10000
> cluster.shd-max-threads: 8
> cluster.locking-scheme: granular
> cluster.data-self-heal-algorithm: full
> cluster.server-quorum-type: server
> cluster.quorum-type: auto
> cluster.eager-lock: enable
> network.remote-dio: enable
> performance.low-prio-threads: 32
> performance.io-cache: off
> performance.read-ahead: off
> performance.quick-read: off
> transport.address-family: inet
> nfs.disable: on
> performance.client-io-threads: off
>
>
> Thanks.
> _______________________________________________
> Users mailing list -- users(a)ovirt.org
> To unsubscribe send an email to users-leave(a)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/HKPAM65CSDJ...
1:52 p.m.
Thank you for the explanation, so from my understanding we can't have a hyperconverged
setup if more that 3 nodes are required? or is it possible if gluster is setup separate of
the oVirt node installs?
2:39 p.m.
From my understanding both are possible. Gluster storage can be completely
separate (at that point you would not be working with a hyperconverged
configuration). You should also be able to expand a hyperconverged setup
past 3 nodes but not by only 1 node, you'd need to add more nodes.
On Thu, Apr 11, 2019 at 8:53 AM <adrianquintero(a)gmail.com> wrote:
Thank you for the explanation, so from my understanding we can't
have a
hyperconverged setup if more that 3 nodes are required? or is it possible
if gluster is setup separate of the oVirt node installs?
_______________________________________________
Users mailing list -- users(a)ovirt.org
To unsubscribe send an email to users-leave(a)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/77URQGHY7LB...
2:49 p.m.
Would you know of any documentation that I could follow for that type of setup?
I have read quite a bit about oVIrt and Hyperconverged but I have only found 3 node setup
examples, nobody seems to go past that or at least I've yet to find one. I already
have a 3 node setup with hyperconverged and working as it should with no issues and tested
fail scenarios, but I need to have an environment with at least 12 servers and be able to
maintain the correct fail scenarios with storage (gluster), however I cant seem to figure
out the proper steps to achieve this.
So if anyone can point me in the right direction it will help a lot, and once I test I
should be able to provide back any knowledge that I gain with such type of setup.
thanks again.
1835
days inactive
1884
days old
4 comments
3 participants
participants (3)
-
adrianquintero＠gmail.com -
Jayme -
Sahina Bose