[Gluster-users] latency limits of glusterfs for replicated mod

[Vlad Kopylov]

Works with same latency for me on 3.12.15
Only problem is that read and write gets slow.
If you host VMs having hypervisor use gfapi: writes will be acceptable, but
read will wonder through nodes, obviously slowing down. 0.3 to 13 ms x 2
around 100 times slower.

There is select local and latency(and queue) read replica selection
functionality in progress https://github.com/gluster/glusterfs/issues/520
Plus halo writes from facehook to make writes replicate on background.
Though I was unable to confirm readiness of any of above functionality in
practice. You can follow or file github issues.

v


On Tue, Nov 13, 2018 at 7:32 AM Oğuz Yarımtepe <oguzyarimtepe at gmail.com>
wrote:

> Hi,
>
> I have 4 node GlusterFS running on two different locations in replicated
> mode. The nodes are splitted equally and the latency is 0.3ms. So i have no
> problem about replicating so far.
>
> Now i want to add two more node at another distance that has |13ms
> latency. Is it said that replication will not be running in a good way and
> thing may get obscured and it is not suggested to use in that much
> latencies.
>
> What is the latency limit for GlusterFS while using replicated mode?
>
> --
> Oğuz Yarımtepe
> http://about.me/oguzy
> _______________________________________________
> Gluster-users mailing list
> Gluster-users at gluster.org
> https://lists.gluster.org/mailman/listinfo/gluster-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.gluster.org/pipermail/gluster-users/attachments/20181113/3763d80e/attachment.html>