apurtell edited a comment on pull request #3244: URL: https://github.com/apache/hbase/pull/3244#issuecomment-836903862
> So we will only compress value? This is an enhancement to existing WAL compression. As you know the existing WAL compression already compresses other aspects of WAL entries _except_ for the value. This patch adds support for compressing values too. > As we will do batching when writing WAL entries out, is it possible to compress when flushing? The data will be larger and compress may perform better. The structure of a WAL file will be multiple compressed blocks. This is not possible for two reasons: 1. WALCellCodec does not compress the WAL file in blocks. The design is edit by edit. I want to introduce value compression without re-engineering the whole WAL format. Perhaps our WAL file format is due for a redesign, but I would like to see that be a different issue. 2. It is not necessary to do that. By using the same Deflater instance for the whole WAL we already get the benefit you are thinking of... it builds its dictionary across the whole file. Even though compression is flushed at every cell, we are not resetting the compressor. (That would be FULL_FLUSH. We are using SYNC_FLUSH.) Way back in the distant past our WAL format was based on Hadoop's SequenceFile, which supported both record-by-record and block based compression, where the blocks would contain multiple records. I don't remember why we moved away from it but I imagine it was because if there are corruptions of the WAL, a record by record codec is able to skip over the corrupt record and we lose only the record (or as many records as are actually corrupt), but with a block format we would lose the whole record and all of the edits contained within that record, especially if compression or encryption is enabled. -- 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]
