ARROW-2551: [Plasma] Improve notification logic

[zhijunfu]

This change targets to improve a few places in current plasma notification code:

1. When a client subscribes to Plasma, the store pushes notifications about existing objects to ALL subscribers, while it should only push to the new subscriber.
2. And in the above scenario, it should only push "sealed" objects to the new subscriber, while currently it pushes all objects regardless of the state.
3. When a client disconnects, it will no longer be able to receive notifications, thus the NotificationQueue for the client should be removed from global map.

[codecov-io]

Codecov Report

Merging #2031 into master will increase coverage by 0.02%.
The diff coverage is 94.73%.

@@            Coverage Diff             @@
##           master    #2031      +/-   ##
==========================================
+ Coverage   86.35%   86.37%   +0.02%     
==========================================
  Files         242      230      -12     
  Lines       41184    40474     -710     
==========================================
- Hits        35565    34960     -605     
+ Misses       5619     5514     -105

Impacted Files	Coverage Δ
cpp/src/plasma/store.h	100% <ø> (ø)	⬆️
cpp/src/plasma/store.cc	88.71% <94.73%> (-0.05%)	⬇️
python/pyarrow/array.pxi	66.04% <0%> (-2.5%)	⬇️
python/pyarrow/feather.pxi	67.44% <0%> (-1.45%)	⬇️
cpp/src/arrow/util/bit-util.h	98.01% <0%> (-1.23%)	⬇️
python/pyarrow/types.pxi	57.9% <0%> (-1.09%)	⬇️
python/pyarrow/table.pxi	68.69% <0%> (-1.07%)	⬇️
python/pyarrow/__init__.py	57.14% <0%> (-0.76%)	⬇️
python/pyarrow/io.pxi	62.23% <0%> (-0.53%)	⬇️
cpp/src/arrow/io/hdfs-internal.cc	33.62% <0%> (-0.45%)	⬇️
... and 52 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 6646864...84f4935. Read the comment docs.

[pcmoritz]

Let's make notification_fd = -1 by default if not set (and get rid of the boost dependency)

[pcmoritz]

This looks good to me! Can you get rid of the boost::option dependency and use the value -1 instead of boost::none? At the moment plasma doesn't depend on boost and I'd like to keep it that way unless there is a good reason to introduce it.

Also can you rebase the PR? There is a small conflict at the moment.

[zhijunfu]

@pcmoritz Thanks for reviewing :) Changed boost::none to -1 and did a rebase to resolve the conflict.

[pcmoritz]

@zhijunfu Sorry, the other PR caused a merge conflict, can you rebase/merge so we can get this PR in? Thanks :)

[zhijunfu]

@pcmoritz Sorry for the long delay in reply, I just rebased the change on top of latest code. The build failures w/ parquet and tool-chain should be unrelated with the change.

[wesm]

Unrelated to the content of this patch: could we regularize the variable name convention in this header and elsewhere in Plasma? Would help with readability

[zhijunfu]

Yes that makes a lot of sense:)

[zhijunfu]

Re-triggered the build and it passed this time.

[wesm]

+1. Merging this, and also opened https://issues.apache.org/jira/browse/ARROW-2690