[
https://issues.apache.org/jira/browse/BEAM-7718?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16883084#comment-16883084
]
Steve Niemitz commented on BEAM-7718:
-------------------------------------
Additionally, the gRPC implementation seems efficient enough that it will very
quickly hit the pubsub subscriber quota with any non-trivial number of workers.
I'm not really sure why, using the windmill-native version I can easily read
from this subscription at a reasonable rate with plenty of quota left.
I wasn't able to do the same with the JSON version, but I think that's just
because it isn't efficient enough to be able to get enough throughput on a
worker.
> PubsubIO to use gRPC API instead of JSON REST API
> -------------------------------------------------
>
> Key: BEAM-7718
> URL: https://issues.apache.org/jira/browse/BEAM-7718
> Project: Beam
> Issue Type: New Feature
> Components: io-java-gcp
> Reporter: Steve Niemitz
> Priority: Major
>
> {quote}The default implementation uses the HTTP REST API, which seems to be
> much less performant than the gRPC implementation. Is there a reason that
> the gRPC implementation is essentially unavailable from the public API?
> PubsubIO.Read.withClientFactory is package private. I worked around this by
> making it public and rebuilding{quote}
--
This message was sent by Atlassian JIRA
(v7.6.14#76016)
