[Gluster-Maintainers] I am seeing nokia folks to be active in the community

Pranith Kumar Karampuri pkarampu at redhat.com
Wed May 25 03:45:06 UTC 2016


This is the reply I got from them: (It is the same Olia who find a leak in
afr data-self-heal of a dict, so I was happy to see the confirmation that
it is indeed the same person :-) )

====================

Hi Pranith,



Yes it’s me! Thank you a lot for your interest! It is a great deal for us!



Actually we are using glusterfs as shared storage solution within VM
clusters running within openstack (in the future it is true, it will be
containers).

Currently our use-case involves glusterfs 3.6.9 utilizing 3 servers with
replication 2 (third server has no bricks).



Our main problems come from the split-brain files and memory usage of
glusterfsd as well as glusterfs fuse clients.

For the first one we are thinking to use replica 3 along with client quorum
(now only server quorum is used).

For the second (mem usage, which is most important), we are trying to
cherry pick commits from 3.7/3.8 series, in order to check if there is any
decrease in the mount of memory server and clients use.

We are trying to upstream them as well for the community in the release
3.6, but smoke builds in BSD have blocked us.



I have also included Vaggelis Polakis in the conversation, who is the
architect in the team and has a better overview of all our issues with
gluster.



We certainly need your help as we are new to how gluster works and any
input from you will be valuable!



Thank you again for your email.

Best Regards,
Olia
====================

I think it will be great if we get these people up to speed. I have seen
how they were helpful without any help from us (They found/fixed 2 memory
leaks). I think it is high time we have enough developer documentation to
help people come up to speed. Let me see if we can get the videos we have
here in Bangalore uploaded in youtube which explain some of the components.
Any other information you guys have that can help them out? Do you guys
think it is better if this discussion happens in gluster-devel mailing list?

Pranith

On Tue, May 24, 2016 at 12:15 PM, Pranith Kumar Karampuri <
pkarampu at redhat.com> wrote:

> Okay, I reached out to both the guys saying hi and how they are using
> gluster. I will let everyone know based on their response.
>
> Pranith
>
> On Tue, May 24, 2016 at 11:52 AM, Raghavendra Gowdappa <
> rgowdapp at redhat.com> wrote:
>
>>
>>
>> ----- Original Message -----
>> > From: "Pranith Kumar Karampuri" <pkarampu at redhat.com>
>> > To: maintainers at gluster.org
>> > Sent: Tuesday, May 24, 2016 11:46:14 AM
>> > Subject: [Gluster-Maintainers] I am seeing nokia folks to be active in
>> the    community
>> >
>> > Anyone knows how they are using gluster? The questions they are asking
>> it
>> > seems like inside containers they are using gluster. Wondering if anyone
>> > else also got mails regarding this to gain more context. We can also
>> > probably ask their usecase just to gain context?
>>
>> Yes. I saw some patches too
>> http://review.gluster.org/#/q/owner:%22Olia+Kremmyda%22+status:open
>>
>> >
>> > --
>> > Pranith
>> >
>> > _______________________________________________
>> > maintainers mailing list
>> > maintainers at gluster.org
>> > http://www.gluster.org/mailman/listinfo/maintainers
>> >
>>
>
>
>
> --
> Pranith
>



-- 
Pranith
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.gluster.org/pipermail/maintainers/attachments/20160525/f942b8f5/attachment.html>


More information about the maintainers mailing list