ivankelly commented on issue #4158: Create more offloader indices per each block uploaded to 2nd tier URL: https://github.com/apache/pulsar/pull/4158#issuecomment-487631515 In my opinion, the issue here is really that we are calling seek each time in the ReadHandle implementation. When I was originally writing this code, my plan was that seek would only occur if you were reading from the stream at an unexpected point. So, if you have read entries 1-100 previously and you came back to the handle asking for 101, it shouldn't seek, as the S3 backed input stream should be in the correct position. If you asked for 102, you would seek. I guess i never actually implemented that though from looking at the code. What are the before and after numbers for this? What is the benchmark setup?
---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
