(Sorry, I missed hdfs-dev@ and my mail filters didn't help either).
As per the email I sent out to general@ a couple of days ago, 0.23.1 is very close right now (there are a handful of perf bugs) after which we should be good for the merge.
On Jan 5, 2012, at 7:52 PM, Eli Collins wrote:
> Hey gang,
> I was looking at the difference between hdfs trunk and 23, for the
> purpose of looking at what to merge to 23.
> Here's a summary of the differences:
> 1. BR scalability (HDFS-395, HDFS-2477, HDFS-2495, HDFS-2476)
> 2. BK support (HDFS-234 and related refactoring)
> 3. Protobuf RPC changes
> 4. A dozen straight-forward bugs and trivial cleanup
> 5. Refactorings (HdfsConstants rename, Refactor RPC out of NN, Untangle NN deps)
> 6. FSEditLog should not writes long and short as UTF8 (HDFS-362)
> 7. Configurable # low resource/failed vols (HDFS-2430)
> 8. Test fixes (for the above)
> Seems like we should merge these:
> #1 - significant improvement, also aids merges of important bug fixes
> like HDFS-1765
> #3 - once it's completed
> #4 - these are no brainers
> #5 - needed to aids future merges like HDFS-1623
> #7 - good feature, not very intrusive
> #8 - required for others
> Seems like we can pass on:
> #2 - is contrib, should soak
> #5 - changes the image format so requires an HDFS upgrade
> I'm planning on merging #4, will update target version on jira before
> doing so, as they should be uncontroversial, but wanted to see what
> Arun and others thought about the other changes. Reasonable?