Skip to content
This repository has been archived by the owner on Dec 19, 2017. It is now read-only.

Latest commit

 

History

History
53 lines (40 loc) · 3.26 KB

CHANGES.md

File metadata and controls

53 lines (40 loc) · 3.26 KB

Change Log

All notable changes to this project will be documented in this file.

Unreleased

1.2.3 - 2017-03-03

Bug Fixes

  • #293 - Relax consistency constraints for client sequencer to ensure missing events cannot prevent the client from progressing

1.2.2 - 2017-02-27

CRITICAL

  • #287 - Fix misaligned lock metadata in segment file headers

Bug Fixes

  • #291 - Ensure server is shut down if cluster leave() fails on leave()
  • #290 - Allow compactor to interrupt compaction threads on shutdown
  • #289 - Complete session unregister if session is already closed
  • #288 - Ensure snapshot is taken in the state machine thread
  • #285 - Too many connection attempts to unavailable followers
  • #284 - Local address is returned as leader by followers in ConnectResponse

1.2.1 - 2017-02-23

CRITICAL

  • #283 - Ensure leader check for linearizable queries occurs after query is applied

Bug Fixes

  • #282 - Reset session command sequence number with keep alive requests
  • #281 - Transport is not properly closed on server leave()
  • #277 - Expire client sessions when command/query fail with unknown session
  • #276 - Simplify management of client connections in servers
  • #271 - Ensure ServerSelectionStrategies.FOLLOWERS can connect to single node

1.2.0 - 2017-01-12

CRITICAL

  • #270 - Remove assert statement with side effects preventing InitializeEntry from being appended
  • #266 and #269 - Prevent commitIndex from being incremented beyond the last index in an AppendRequest

Bug Fixes

  • #267 - Limit request queue size to prevent stack overflow when queue is drained
  • #258 - Fix improper session removal in certain cases after client recovers a session
  • #249 - Fix incorrect exception arguments in state machine exceptions
  • #247 - Fix 100% CPU usage due to compaction thread count
  • #244 - Ensure interval is properly used in state machine scheduler
  • #240 - Fix client state change/recovery bugs
  • #238 - Ensure client can be closed when a majority of the cluster is down

Features

  • #263 - Support client-provided per-session timeouts
  • #262 - Support user-provided client IDs