Michael Han

sure, this was on my list anyway.

>> there will be lots of txns need to be replayed between NEWLEADER and UPTODATE packet. I think these are the transactions queued on learner while SNAP sync is happening?...

>> Those are the txns queued on leader while learner took snapshot when received NEWLEADER packet. Are these txns queued in `queuedPackets` of `LearnerHandler`? I was thinking you referred to...

>> In general, this won't affect the quorum up time, but it may take longer time for the minority server who is lagging behind too much. One case I was...

+1 on the proposal.

Convert this PR to draft status as it's not ready to merge (thanks @ctubbsii for the pointer).

@narendly thank you for feedback. >> Generifying components such as the storage provider implementation yes a more generalized storage provider interface is of my interest as well. something like org.apache.hadoop.fs.FileSystem...

>> This is to ensure that backup continues on in the case of the backup node failure. Currently we just rely on monitoring + alerts and let on call to...

patch LGTM; synchronized on `QuorumPeer` when accessing `shuttingDownLE` is easy to reason and understand over other solutions like making `shuttingDownLE` volatile, which might work but is not so obvious and...

General comment: please add more description on the pull request on what's the motivation of change and how it impacts performance / improve debugging etc. Or add those to the...