when pinned priority policy plan to release?

[z0mb1ek]

Hi! I see in code

// TODO - PINNED CONSUMER SUPPORT
//    PinnedClient("pinned_client")

when do you plan to release it?

[scottf]

This is currently being delayed by the server team. I do not have an ETA

[z0mb1ek]

Am I right in understanding that I can achieve this through prioritization? by setting large unattainable parameters?

[scottf]

I see what you are getting at, but I don't think so. Priority is just an ordinal, and the server decides when to switch. You would start all the consumers at the same time giving priority, and when # 1 goes away, the server switches to # 2.
Pinning is controlled by the user at the time of the consumption. You would start a consume and pin it. But then later on you could start another consume instance and pin that instance and the server will switch.

[z0mb1ek]

That's clearer, thanks.

[z0mb1ek]

and one more question: why there is no batch fetch for priority functional? i need to do pull with PullRequestOptions and then while cycle for nextMessage is not null, It's not very convenient

[z0mb1ek]

it is very sad that you are closing the topic because it is not at all clear if this is a bug or a feature and whether it can be used.

[scottf]

There is not a bug, but it may not make sense to use priority with fetch, I will follow up separately. Please review the examples to try to understand the simplification or not. If you want help with a specific use case and application architecture, Synadia is available for paid support and consulting.

[z0mb1ek]

Its work with prioritization if i make .maxMessages(Int.MAX_VALUE), can i use this in production?

[z0mb1ek]

And i get sense for this making partitioning like kafka with prioritizing consumers

[z0mb1ek]

and there is no failover at all

[scottf]

@z0mb1ek #1472