How can I set a default node?

[AltOscar]

Hello everyone!

I'm running some tests on gluster and I have a couple of questions.

Gluster version is 9.5

I'll give you some background, we have two environments, A and B.
A is located in Europe and B is located in America. I know there'll be issues related to ping etc but for the time being, let's ignore that.
The conf is basic and we want to set it as a replicated-distributed volume.

We'll have clients both on Europe and America but we want both of them bot read files from A("main" node). Of course, if by some chance A loses connection or what not, it'll have to read from B.
We also know about Geo-replication but for now we want to focus on this setup.

So, the questions are as follows:

First, as title says, is there a way to set a default node from which it'll read? So far the tests we've been running seem random

Second, what happens if we have different read/write speeds on each node? (for this question, let's asume both nodes are on LAN, so just asking about pure disk speed). The speed will be the same as the lowest one? Will it write on the fastest one, finish, and still write on the slowest? For now we can't really test this, so if someone has an answer I'd apreciatte it.

Thank you all and if you need more info feel free to ask!

Also English ain't my mother tongue so I'm sorry if there are any mistakes (also apreciatte help here!)

[pranithk]

Not sure if you will get a lot of responses, this is not something people do.

[mykaul]

Wouldn't latency based read child policy help here?

[pranithk]

Lookups will still go to both bricks. So depends on the workload. If the file I/O is dominated by reads/writes then maybe it will work. But if it is dominated by lookups, then latency effects will be faced by the application/user.

[AltOscar]

Sorry for the late response! Thank you so much for your answers, as you said this doesn't seem the norm at all