tarmacmonsterg commented on issue #24380:
URL: https://github.com/apache/pulsar/issues/24380#issuecomment-2949226666

   > 
   > Thanks for sharing, [@tarmacmonsterg](https://github.com/tarmacmonsterg). 
Just curious, how did you address the problem?
   > 
   
   Initially, we had delayed topics split by domain. In some of these topics, 
there were messages scheduled for delivery in 2–3 months, which blocked ledger 
cleanup and triggered a full re-read of all messages stored in the ledgers 
after broker restart.
   
   To prevent this behavior, we reorganized the messages into topics based on 
their delivery time:
        •       10 minutes
        •       1 hour
        •       1 day
        •       1 month
        •       infinity
   
   Since most of our messages fall within the 10-minute range, cleanup now 
occurs in a timely manner, and restarts no longer lead to re-reading large 
volumes of messages.
   
   However, a restart still resets the cursor state, requiring already-read 
messages to be read again. To handle this, the application skips duplicates.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to