[Gluster-users] thin arbiter vs standard arbiter

[wkmail]

I see mentions of thin arbiter in the 4.x notes and I am intrigued.

As I understand it, the thin arbiter volume is

a) receives its data on an async basis (thus it can be on a slower
link). Thus gluster isn't waiting around to verify if it actually got
the data.

b) is only consulted in situations where Gluster needs that third vote,
otherwise it is not consulted.

c) Performance should therefore be better because Gluster is only
seriously talking to 2 nodes instead of 3 nodes (as in normal arbiter or
rep 3)

Am I correct?

If so, is thin arbiter ready for production or at least use on
non-critical workloads?

How safe is it for VMs images (and/or VMs with sharding)?

How much faster is thin arbiter setup over a normal arbiter given that
the normal data only really sees the metadata?

In a degraded situation (i.e. loss of one real node), would having a
thin arbiter on a slow link be problematic until everything is healed
and returned to normal?

Sincerely,

-wk

[Amar Tumballi]

This recently added document talks about some of the technicalities of the
feature:

https://docs.gluster.org/en/latest/Administrator%20Guide/Thin-Arbiter-Volumes/

Please go through and see if it answers your questions.

-Amar

Post by wkmail
I see mentions of thin arbiter in the 4.x notes and I am intrigued.
As I understand it, the thin arbiter volume is
a) receives its data on an async basis (thus it can be on a slower link).
Thus gluster isn't waiting around to verify if it actually got the data.
b) is only consulted in situations where Gluster needs that third vote,
otherwise it is not consulted.
c) Performance should therefore be better because Gluster is only
seriously talking to 2 nodes instead of 3 nodes (as in normal arbiter or
rep 3)
Am I correct?
If so, is thin arbiter ready for production or at least use on
non-critical workloads?
How safe is it for VMs images (and/or VMs with sharding)?
How much faster is thin arbiter setup over a normal arbiter given that the
normal data only really sees the metadata?
In a degraded situation (i.e. loss of one real node), would having a thin
arbiter on a slow link be problematic until everything is healed and
returned to normal?
Sincerely,
-wk
_______________________________________________
Gluster-users mailing list
https://lists.gluster.org/mailman/listinfo/gluster-users

[W Kern]

Post by Amar Tumballi
This recently added document talks about some of the technicalities of
https://docs.gluster.org/en/latest/Administrator%20Guide/Thin-Arbiter-Volumes/
Please go through and see if it answers your questions.
-Amar

Well yes that does answer some. By skipping a lot more of the arbiter
traffic, there may be some noticeable performance benefits especially in
an older 1G network.
At least until you have to deal with a failure situation.

Though the "would you use it on a VM, either now or when the code is
more seasoned?" question is still there.

I'm willing to try it out on some non-critical VMs (cloud-native stuff,
where I always spawn from a golden image), but if it is not ready for
production, then I don't want to bother with it at the moment.

-wk

Post by Amar Tumballi
I see mentions of thin arbiter in the 4.x notes and I am intrigued.
As I understand it, the thin arbiter volume is
a) receives its data on an async basis (thus it can be on a slower
link). Thus gluster isn't waiting around to verify if it actually
got the data.
b) is only consulted in situations where Gluster needs that third
vote, otherwise it is not consulted.
c) Performance should therefore be better because Gluster is only
seriously talking to 2 nodes instead of 3 nodes (as in normal
arbiter or rep 3)
Am I correct?
If so, is thin arbiter ready for production or at least use on
non-critical workloads?
How safe is it for VMs images (and/or VMs with sharding)?
How much faster is thin arbiter setup over a normal arbiter given
that the normal data only really sees the metadata?
In a degraded situation (i.e. loss of one real node), would having
a thin arbiter on a slow link be problematic until everything is
healed and returned to normal?
Sincerely,
-wk
_______________________________________________
Gluster-users mailing list
https://lists.gluster.org/mailman/listinfo/gluster-users
<https://lists.gluster.org/mailman/listinfo/gluster-users>
--
Amar Tumballi (amarts)

[Ravishankar N]

Post by W Kern

Post by Amar Tumballi
This recently added document talks about some of the technicalities
https://docs.gluster.org/en/latest/Administrator%20Guide/Thin-Arbiter-Volumes/
Please go through and see if it answers your questions.
-Amar

Well yes that does answer some. By skipping a lot more of the arbiter
traffic, there may be some noticeable performance benefits especially
in an older 1G network.
At least until you have to deal with a failure situation.
Though the "would you use it on a VM, either now or when the code is
more seasoned?" question is still there.
I'm willing to try it out on some non-critical VMs (cloud-native
stuff, where I always spawn from a golden image), but if it is not
ready for production, then I don't want to bother with it at the moment.

Hi WK,

There are a few patches [1]  that are still undergoing review .  It
would be good to wait for some more time until trying it out. If you are
interested in testing, I'll be happy to inform you once they get merged.

[1] https://review.gluster.org/#/c/20095/,
https://review.gluster.org/#/c/20103/, https://review.gluster.org/#/c/20577/

Regards,
Ravi

Post by W Kern
-wk

Post by Amar Tumballi
I see mentions of thin arbiter in the 4.x notes and I am intrigued.
As I understand it, the thin arbiter volume is
a) receives its data on an async basis (thus it can be on a
slower link). Thus gluster isn't waiting around to verify if it
actually got the data.
b) is only consulted in situations where Gluster needs that third
vote, otherwise it is not consulted.
c) Performance should therefore be better because Gluster is only
seriously talking to 2 nodes instead of 3 nodes (as in normal
arbiter or rep 3)
Am I correct?
If so, is thin arbiter ready for production or at least use on
non-critical workloads?
How safe is it for VMs images (and/or VMs with sharding)?
How much faster is thin arbiter setup over a normal arbiter given
that the normal data only really sees the metadata?
In a degraded situation (i.e. loss of one real node), would
having a thin arbiter on a slow link be problematic until
everything is healed and returned to normal?
Sincerely,
-wk
_______________________________________________
Gluster-users mailing list
https://lists.gluster.org/mailman/listinfo/gluster-users
<https://lists.gluster.org/mailman/listinfo/gluster-users>
--
Amar Tumballi (amarts)

_______________________________________________
Gluster-users mailing list
https://lists.gluster.org/mailman/listinfo/gluster-users

[WK Lists]

Hi WK,

There are a few patches [1]  that are still undergoing review . It
would be good to wait for some more time until trying it out. If you
are interested in testing, I'll be happy to inform you once they get
merged.
[1] https://review.gluster.org/#/c/20095/,
https://review.gluster.org/#/c/20103/,
https://review.gluster.org/#/c/20577/
Regards,
Ravi

yes please let me know when you think the thin-arbiter is "testing" ready.

Again, I have some VM environments that can handle a storage disaster
(though it would be annoying)

[Dmitry Melekhov]

Post by Amar Tumballi
This recently added document talks about some of the technicalities of
https://docs.gluster.org/en/latest/Administrator%20Guide/Thin-Arbiter-Volumes/
Please go through and see if it answers your questions.
-Amar

Hello!

I have question:

Manual says:


"When one brick is up: Fail FOP with EIO."

So, if we have 2 nodes with thin arbiter and only one node is up, i.e.
second node is down for some reason, then I/O will be stopped.
Any reasons to have two nodes then?

Could you tell me is manual right here or it is misprint?

Thank you!

Post by Amar Tumballi
I see mentions of thin arbiter in the 4.x notes and I am intrigued.
As I understand it, the thin arbiter volume is
a) receives its data on an async basis (thus it can be on a slower
link). Thus gluster isn't waiting around to verify if it actually
got the data.
b) is only consulted in situations where Gluster needs that third
vote, otherwise it is not consulted.
c) Performance should therefore be better because Gluster is only
seriously talking to 2 nodes instead of 3 nodes (as in normal
arbiter or rep 3)
Am I correct?
If so, is thin arbiter ready for production or at least use on
non-critical workloads?
How safe is it for VMs images (and/or VMs with sharding)?
How much faster is thin arbiter setup over a normal arbiter given
that the normal data only really sees the metadata?
In a degraded situation (i.e. loss of one real node), would having
a thin arbiter on a slow link be problematic until everything is
healed and returned to normal?
Sincerely,
-wk
_______________________________________________
Gluster-users mailing list
https://lists.gluster.org/mailman/listinfo/gluster-users
<https://lists.gluster.org/mailman/listinfo/gluster-users>
--
Amar Tumballi (amarts)
_______________________________________________
Gluster-users mailing list
https://lists.gluster.org/mailman/listinfo/gluster-users

[Ashish Pandey]

I think it should be rephrased a little bit -

"When one brick is up: Fail FOP with EIO."
should be
"When only one brick is up out of 3 bricks: Fail FOP with EIO."

So we have 2 data bricks and one thin arbiter brick. Out of these 3 bricks if only one brick is UP then we will fail IO.

---
Ashish


----- Original Message -----

From: "Dmitry Melekhov" <***@belkam.com>
To: gluster-***@gluster.org, ***@redhat.com
Sent: Thursday, August 2, 2018 4:59:41 PM
Subject: Re: [Gluster-users] thin arbiter vs standard arbiter

01.08.2018 22:04, Amar Tumballi пОшет:



This recently added document talks about some of the technicalities of the feature:

https://docs.gluster.org/en/latest/Administrator%20Guide/Thin-Arbiter-Volumes/

Please go through and see if it answers your questions.

-Amar



Hello!

I have question:

Manual says:


"When one brick is up: Fail FOP with EIO."

So, if we have 2 nodes with thin arbiter and only one node is up, i.e. second node is down for some reason, then I/O will be stopped.
Any reasons to have two nodes then?

Could you tell me is manual right here or it is misprint?

Thank you!



<blockquote>



On Wed, Aug 1, 2018 at 11:09 PM, wkmail < ***@bneit.com > wrote:

<blockquote>
I see mentions of thin arbiter in the 4.x notes and I am intrigued.

As I understand it, the thin arbiter volume is

a) receives its data on an async basis (thus it can be on a slower link). Thus gluster isn't waiting around to verify if it actually got the data.

b) is only consulted in situations where Gluster needs that third vote, otherwise it is not consulted.

c) Performance should therefore be better because Gluster is only seriously talking to 2 nodes instead of 3 nodes (as in normal arbiter or rep 3)

Am I correct?

If so, is thin arbiter ready for production or at least use on non-critical workloads?

How safe is it for VMs images (and/or VMs with sharding)?

How much faster is thin arbiter setup over a normal arbiter given that the normal data only really sees the metadata?

In a degraded situation (i.e. loss of one real node), would having a thin arbiter on a slow link be problematic until everything is healed and returned to normal?

Sincerely,

-wk

_______________________________________________
Gluster-users mailing list
Gluster-***@gluster.org
https://lists.gluster.org/mailman/listinfo/gluster-users



</blockquote>
--
Amar Tumballi (amarts)


_______________________________________________
Gluster-users mailing list Gluster-***@gluster.org https://lists.gluster.org/mailman/listinfo/gluster-users

</blockquote>




_______________________________________________
Gluster-users mailing list
Gluster-***@gluster.org
https://lists.gluster.org/mailman/listinfo/gluster-users

[Dmitry Melekhov]

Post by Ashish Pandey
I think it should be rephrased a little bit -
"When one brick is up: Fail FOP with EIO."
should be
"When only one brick is up out of 3 bricks: Fail FOP with EIO."
So we have 2 data bricks and one thin arbiter brick. Out of these 3
bricks if only one brick is UP then we will fail IO.
---
Ashish

Hello!

Thank you!

This is what we need :-)

Post by Ashish Pandey
------------------------------------------------------------------------
*Sent: *Thursday, August 2, 2018 4:59:41 PM
*Subject: *Re: [Gluster-users] thin arbiter vs standard arbiter
This recently added document talks about some of the
https://docs.gluster.org/en/latest/Administrator%20Guide/Thin-Arbiter-Volumes/
Please go through and see if it answers your questions.
-Amar
Hello!
"When one brick is up: Fail FOP with EIO."
So, if we have 2 nodes with thin arbiter and only one node is up, i.e.
second node is down for some reason, then I/O will be stopped.
Any reasons to have two nodes then?
Could you tell me is manual right here or it is misprint?
Thank you!
I see mentions of thin arbiter in the 4.x notes and I am intrigued.
As I understand it, the thin arbiter volume is
a) receives its data on an async basis (thus it can be on a
slower link). Thus gluster isn't waiting around to verify if
it actually got the data.
b) is only consulted in situations where Gluster needs that
third vote, otherwise it is not consulted.
c) Performance should therefore be better because Gluster is
only seriously talking to 2 nodes instead of 3 nodes (as in
normal arbiter or rep 3)
Am I correct?
If so, is thin arbiter ready for production or at least use on
non-critical workloads?
How safe is it for VMs images (and/or VMs with sharding)?
How much faster is thin arbiter setup over a normal arbiter
given that the normal data only really sees the metadata?
In a degraded situation (i.e. loss of one real node), would
having a thin arbiter on a slow link be problematic until
everything is healed and returned to normal?
Sincerely,
-wk
_______________________________________________
Gluster-users mailing list
https://lists.gluster.org/mailman/listinfo/gluster-users
--
Amar Tumballi (amarts)
_______________________________________________
Gluster-users mailing list
https://lists.gluster.org/mailman/listinfo/gluster-users
_______________________________________________
Gluster-users mailing list
https://lists.gluster.org/mailman/listinfo/gluster-users