rmahindra123 opened a new pull request #3128: URL: https://github.com/apache/hudi/pull/3128
## What is the purpose of the pull request Based on our benchmark of ExternalSpillableMap with large data, we noticed that higher amount of data is spilled to disk when using DiskBasedMap when compared to RocksDb, causing much higher GET latencies. One strong reason for this was that compression was not being used in DiskBasedMap, while RocksDb uses compression. This PR adds a way to enable compression when storing KVs in DiskBasedMap that should alleviate such perf concerns. ## Brief change log - Added compression and decompression when storing KVs in DiskBasedMap.java ## Verify this pull request - Added tests to TestDiskBasedMap.java and TestExternalSpillableMap.java -- 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]
