List of issues addressed in the 3.1.4 release
State of vBucket on new master set to "replica" instead of "pending" during DCP takeover
Revert change to set state of vbucket to "pending" (instead of replica) on new master during takeover (MB-16292)
Rebalance performance regressed in 3.1.4
XDCR Replication regressed in 3.1.4
View query latency slower in 3.1.4-1836 (compared to 1835).
Manual actions after Upgrade
Wrong high sequence number returned from ep-engine
[Backport 3.1.x] Check that the file handle on the memcached log was open correctly
Janitor activation of vbucket post rebalance failure results in unnoticed dead DCP stream
mcd abort during rebalance
CLONE - [revAB] 80% regression in APPEND operation latency
A race between compaction thread and the memcached thread that changes the vbucket state could fire an assertion crash
Reduce lock contention between store operations and DCP
DCP Producer could miss streaming items from certain streams
Can't retrieve sequence number
Explicitly close DCP streams after abrupt rebalance termination
[system test] Rebalance failed with 'setup_replications_failed'
ep-engine: circular reference between ActiveStream & ActiveStreamCheckpointProcessorTask causes memory leak.
Generate new UUID for a node when it is added to a cluster
Log separate messages for checkpoint persistence and seqno persistence commands
EP engine should log a message and relevant stats when it changes the state of vBucket to dead