captainzmc edited a comment on pull request #2495: URL: https://github.com/apache/ozone/pull/2495#issuecomment-894114695
> * Let's do not extend OutputStream in all the new classes in order to keep the code simple. If there is a need, we may add it later. > * In BlockDataStreamOutput, we should write directly to the underlying stream. We should not use BufferPool and ChunkBuffer in BlockDataStreamOutput anymore since we want to avoid buffer copying. Hi @szetszwo, I have a question about compatibility here. We are using HDFS APi, using ofs:// or o3fs:// to access Ozone. The old app used like this: `FSDataOutputStream fsOut = fs.create(new Path(pathStr), true);` `fsOut.write(str.getBytes());` Old applications, such as MapReduce or Spark Tasks, pass in a byte array, which is difficult to change. So can we still keep the OutputStream? Because we need its write(byte b[]). Although using this API will increase the buffer copy of the client, it will resolve compatibility issues. -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
