Gluster users - Dec 2020 - Replication logic

Hello all

I'm new to gluster and to this list too and, as is to be expected,
I come with questions.

I have set up a replica 3 arbiter 1 volume. Is there a way to turn
the arbiter into a full replica without breaking the volume and
losing the metadata that is already on the arbiter?

Given a normal (not geo) replica 3 volume like this

      brick1
     /      \
    /        \
 brick2-----brick3

where brick2<->brick3 is a high-speed connection, but
brick1<->brick2
and brick1<->brick3 are low speed, and data is fed to brick1, is there
a way to tell the volume that brick1 should only feed brick2 and let
brick2 feed brick3 if (and only if) all three are online, rather than
brick1 feeding both brick2 and brick3?

I realise that taking brick3 offline during a data burst and then back
online would approximate this effect, but my question is about a
permanent arrangement, something like priorities or weights.

If there is no way to configure this, does brick1 in the default
behaviour send all files sequentially to both other bricks, or does
it initially send different files to each of them, so that they get
a chance to also share between themselves?

Z

Merry Christmas!
>I have set up a replica 3 arbiter 1 volume. Is there a way to turn
>the arbiter into a full replica without breaking the volume and
>losing the metadata that is already on the arbiter?
Yes, you have to use "remove-brick" with the option
"replica" to reduce the replica count and then reformat the arbiter
brick and add it back.

>where brick2<->brick3 is a high-speed connection, but
brick1<->brick2
>and brick1<->brick3 are low speed, and data is fed to brick1, is there
>a way to tell the volume that brick1 should only feed brick2 and let
>brick2 feed brick3 if (and only if) all three are online, rather than
>brick1 feeding both brick2 and brick3?
Erm... this is not how it works. The FUSE client (mount -t glusterfs) is writing
to all bricks in the replica volume, not the brick to brick.What you have
described is the old legacy NFS (currently requires rebuild from source).

>If there is no way to configure this, does brick1 in the default
>behaviour send all files sequentially to both other bricks, or does
>it initially send different files to each of them, so that they get
>a chance to also share between themselves?
As I mentioned , the FUSE client is replicating the data during writes, while
uses the fastest responding brick for reads.

What are you trying to achive ? What is your setup (clients,servers,etc) ?

Best Regards,
Strahil Nikolov