sodonnel commented on PR #3286: URL: https://github.com/apache/ozone/pull/3286#issuecomment-1091905512
With the approach suggested here, if we no longer check quota when allocating blocks, I could write a very large key only to find it to beyond the quota when I commit it - this is potentially gigabytes of data. If I am writing something like streaming data, it may be hard to "rewind" the stream to write it all again. It is more user friendly IMO, that when you get a block allocated, its guaranteed to not break the quota as your space is pre-allocated. If the issue is mainly around dropping a bucket, can we do anything at bucket drop time? Ideally we should not be able to drop a bucket that still contains keys, so could the drop bucket operation check the open key table and manage its open keys and quota accordingly before the drop bucket completes? -- 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]
