You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
#968 made transactions optional. But there is code underneath it that assumes it happens in a transaction and discards a future from the producer client:
// we intentionally suppress FutureReturnValueIgnored here in errorprone - this is because// we wrap this in a transaction, which is responsible for flushing all of the pending// messageskafkaProducer.send(producerRecord);
If the producer has an error when the transactions are disabled, it won't get surfaced in logs or http responses. Instead, it will cause requests to the bulk ingest endpoint to silently time out.
Requirements (place an x in each of the [ ])**
I've read and understood the Contributing guidelines and have done my best effort to follow them.
I've searched for any related issues and avoided creating a duplicate issue.
To Reproduce
Steps to reproduce the behavior:
One option would be to misconfigure kafka with transactions disabled. Eg, a dataset with a partition listed that doesn't exist, or bad auth or a unreasonably short timeout.
But you can also see it by changing the transaction system property setting in the BulkIngestKafkaProducerTest to false
A clear and concise description of what you expected to happen.
Errors are logged
Screenshots
If applicable, add screenshots to help explain your problem.
Reproducible in:
Astra version: HEAD
JVM version:
OS version(s):
Additional context
Add any other context about the problem here.
Related: #618 - The misconfiguration we had that triggered this is related to this other issue, but it is separate in that the logs having no information instead of "a significant amount of confusing messages".
The text was updated successfully, but these errors were encountered:
Describe the bug
#968 made transactions optional. But there is code underneath it that assumes it happens in a transaction and discards a future from the producer client:
https://github.com/slackhq/astra/blob/master/astra/src/main/java/com/slack/astra/bulkIngestApi/BulkIngestKafkaProducer.java#L336-L339
If the producer has an error when the transactions are disabled, it won't get surfaced in logs or http responses. Instead, it will cause requests to the bulk ingest endpoint to silently time out.
Requirements (place an
x
in each of the[ ]
)**To Reproduce
Steps to reproduce the behavior:
One option would be to misconfigure kafka with transactions disabled. Eg, a dataset with a partition listed that doesn't exist, or bad auth or a unreasonably short timeout.
But you can also see it by changing the transaction system property setting in the BulkIngestKafkaProducerTest to false
Expected behavior
A clear and concise description of what you expected to happen.
Errors are logged
Screenshots
If applicable, add screenshots to help explain your problem.
Reproducible in:
Astra version: HEAD
JVM version:
OS version(s):
Additional context
Add any other context about the problem here.
Related: #618 - The misconfiguration we had that triggered this is related to this other issue, but it is separate in that the logs having no information instead of "a significant amount of confusing messages".
The text was updated successfully, but these errors were encountered: