Export - CSV (All fields)
Export - CSV (Current fields)
KUDU-1883
Support Ruby 2.4 in make docs taskKUDU-105
Support for timeouts on client admin ops
Demote or remove WARNING about large arenas
block cache limit seems to not be fully respectedKUDU-579
Scanner fault toleranceKUDU-463
Add checksumming to cfile and other on-disk formats
Java client cleanup for release
Massive numbers of threads used by log append and GC
Create an integration test that performs writes with multiple consistency modes
[java client] Refactor RPC to avoid the "master hack"
add an environment variable for kudu client debugging to stderr
Master UI does not list replicas that are part of a "pending" config
compaction/flush logs should include tablet ID
Make ksck support safe time and rework ksck snapshot tests
Java client doesn't support fault tolerant scans
disk size is much large than actually data size
Unable to delete FAILED tablets as part of drop table
The validation of 'grouped' flags works incorrectly if flags re-orderedKUDU-2021
Add an integration test: in case of multi-master, RPC is retried in case of negotiation error with master
Spark read fails if key columns are not leading columns
KUDU-1917
Add tests which check for KDC fault conditions
Add client request sidecars to KRPC
Kudu C++ client: re-acquire a new authn token when current one expires
Master does not retry TS maintenance task if target TS not registered
Incorrect calculation determining performance improvement for flush
Server should refuse to start if authentication is 'required' but no auth method is configured
Improve performance of dictionary builder
ksck doesn't identify tablets that are evicted but still in config
Document Kudu command-line tools
Reduce Kudu WAL log disk usage
timestamp propagation APIs missing from synchronous Java client
java client: Impossible to uniquely identify error rows because PartialRow impl has no getters
Replicated master process (no master SPOF)
Client does not fail over due to timeout
MemTracker root tracker consumption is rarely updated
Java client logs NPE if a connection is closed by client during negotiation
MM should trigger flushes before memory backpressure kicks in
Cannot create Kudu table with Range Partitioning
Java openTable() call eagerly connects to all tablets
OpId index 32-bit overflow (was: Master crashes after too many TS re-registrations)
Improve tablet disk space estimation
Error during tablet copy may orphan a bunch of stuff
Connection negotiation timeout to tablet server is treated as unretriable error on write operations
heap-use-after-free when running threadlocal cache destructor
NPE and lost callback in Java client when master queue overflows