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
Quickwit ensures exactly-once semantics by storing Kafka offsets in the metastore on a split publish.
We may also want to commit offsets after a split publish on Kafka side too. The benefit would be that Kafka will be aware of the lag of the consumers, it can be helpful for monitoring purposes.
However, I'm not sure if we can do it correctly, we would need to commit offsets on Kafka without interfering with our Kafka consumers.
There is also an alternative: a monitoring dashboard could compare the latest publish offsets by calling the Quickwit API and thus be available to track the lag.
The text was updated successfully, but these errors were encountered:
Quickwit ensures exactly-once semantics by storing Kafka offsets in the metastore on a split publish.
We may also want to commit offsets after a split publish on Kafka side too. The benefit would be that Kafka will be aware of the lag of the consumers, it can be helpful for monitoring purposes.
However, I'm not sure if we can do it correctly, we would need to commit offsets on Kafka without interfering with our Kafka consumers.
There is also an alternative: a monitoring dashboard could compare the latest publish offsets by calling the Quickwit API and thus be available to track the lag.
The text was updated successfully, but these errors were encountered: