Skip to content

Latest commit

 

History

History
107 lines (82 loc) · 2.82 KB

CHANGELOG.md

File metadata and controls

107 lines (82 loc) · 2.82 KB

Overlord-proxy

Version 1.9.0

  1. complete memcache binary protocol.
  2. fix memcache text request key&data no copy bug.

Version 1.8.5

  1. add slowlog file rotate based on size
  2. add slowlog file total limit
  3. increasing node connection inputs size

Version 1.8.4

  1. fix time record MarkEndInput, which reduce memory consumption by 60%
  2. try fetch cluster nodes when key moved

Version 1.8.3

  1. memcache/protocol: impl quit and noreply commands
  2. add RequestTypeVersion command for java xmemcache client
  3. refactor proxy connection for overlord
  4. support sunionstore and zunionstore

Version 1.8.2

  1. add overlord version of all sub commands
  2. add anzi rdb callback ignore receive for io.discard
  3. fixed reload bugs when add/reduce server
  4. add read sleep for flushing disk

Version 1.8.0

  1. add memcache proxy support and fixed reset subs
  2. add slowlog basic functions
  3. fixed anzi bugs and add more logging

Version 1.7.4

  1. reduce node connection default memory allocation to 512k.
  2. fix memcache node connection bugs
  3. reduce prometheus memory usage when handles connection errors.

Version 1.7.1

  1. set network io err to unify error in case of mem boom

Version 1.7.0

  1. add reload feature for cache type: redis, memcache, memcache_binary
  2. add more metrics
  3. add fuzz test for redis/memcache protocol

Version 1.6.0

  1. add migrate tools named anzi
  2. add cluster manager tools enri
  3. fixed overlord panic when all cluster seed nodes down
  4. add back prometheus monitor metrics

Version 1.5.5

  1. change pinger as long connection.

Version 1.5.4

  1. ping timeout 100ms.
  2. fix max connection limit.

Version 1.5.3

  1. fix pipe read when one err.

Version 1.5.2

  1. max redirects 5.

Version 1.5.1

  1. reset sub message only in need.

Version 1.5.0

  1. refactor message pipeline.
  2. non-persistent connection for cluster redirect.

Version 1.4.0

  1. add redis cluster support.

Version 1.3.2

  1. change round-chan to race-chan.

Version 1.3.1

  1. hot fix reconn.

Version 1.3.0

  1. add executor interface.
  2. add pinger interface.
  3. add MsgBatchAlloctor for one node WR.

Version 1.2.2

  1. fix batchdone err.
  2. add node reconn.

Version 1.2.1

  1. fix message and request reuse bug.

Version 1.2.0

  1. add redis protocol support.

Version 1.1.0

  1. add memcache binary protocol support.
  2. add conf file check.

Version 1.0.0

  1. compitable consist hash with twemproxy.
  2. reduce object alloc (by using Pool).
  3. recycle using of request and response memory.
  4. batch message execute to reduce conflict of command channel.
  5. using writev to reduce syscall times.
  6. synchronously send and wait of message.
  7. zero copy of request and response data at all.