Release 1.11.6

[neverchanje]

This is a minor release, including some bugfixes.

• utils: fix a bug in table_printer and add several unit tests XiaoMi/rdsn#265 (XiaoMi/rdsn@83959da)

• replication: fix cold backup bug & config loading XiaoMi/rdsn#267 (XiaoMi/rdsn@e726a81)

• bugfix: fix error handling in filesystem::file_tree_walk XiaoMi/rdsn#268 (XiaoMi/rdsn@314cd90)

• scripts: build rocksdb with RelWithDebInfo #355 (bfc3421)

Scripts and docs:

• docker: add scripts for packing docker image & local docker cluster d… #335 (040af1e)

• scripts: add convenience script for creating table #348 (2887873)

• docs: update readme of Pegasus #349 (547f647)

Minor fixes:

• remote-cmd: make HANDLE_CLI_FLAGS a short line of macro XiaoMi/rdsn#229 (XiaoMi/rdsn@b464267)

• config: add configuration duplication_disabled XiaoMi/rdsn#230 (XiaoMi/rdsn@a493678)

• utils: test format utilities XiaoMi/rdsn#231 (XiaoMi/rdsn@fc0f324)

• log: reduce some rocksdb statistics log #326 (5cf8b7a)

• refactor: remove some useless configs #338 (3c59ffd)

• *: improve the method of modify the config in test #352 (7ff7cd3)

Configuration updates:

[core]
- ;tool = fastrun
- cli_local = false
- cli_remote = false
- start_nfs = false
- ;logging_factory_name = dsn::tools::hpc_logger

- [tools.hpc_logger]
- per_thread_buffer_bytes = 8192

[threadpool.THREAD_POOL_DEFAULT]
- max_input_queue_length = 1024

[replication]
+ duplication_disabled = true

[pegasus.server]
- updating_rocksdb_sstsize_interval_seconds = 600
+ update_rdb_stat_interval = 600

[task..default]
- fast_execution_in_network_thread = false

[neverchanje]

The above commits have been cherry-picked in 1.11.6-RC1.
Additional commits to be added to 1.11.6-RC2:

Important bugfix on failure detection!

• fd: fix ingeter overflow XiaoMi/rdsn#272 (XiaoMi/rdsn@75029e0)

• fd: fix unit test XiaoMi/rdsn#279

• fd: improve failure detector to avoid getting now time out of lock XiaoMi/rdsn#277

Whitelist support:

• add unit test of whitelist & fix bug XiaoMi/rdsn#228 (XiaoMi/rdsn@f9bb246)

• meta: support white list of replica-servers & add remote cmd to ctrl live_percentage XiaoMi/rdsn#226 (XiaoMi/rdsn@ddaa1ad)

Server connection threshold support:

• network: server connection threshold XiaoMi/rdsn#234 (XiaoMi/rdsn@ca636f2)

Minor fixes:

• *:fix Semaphore bug & Compile warnings XiaoMi/rdsn#238 (XiaoMi/rdsn@0048ae4)

• rpc: support sending rpc_holder through partition_resolver XiaoMi/rdsn#227 (XiaoMi/rdsn@36607b7)

• rpc: minor refactoring on thrift_messsage_parser XiaoMi/rdsn#251 (XiaoMi/rdsn@3989922)

• rpc: minor refactoring on message_reader XiaoMi/rdsn#252 (XiaoMi/rdsn@270a42c)

Periodically memory release:

• utils: periodically memory release of tcmalloc XiaoMi/rdsn#278

Redis Proxy:

• GEO: Support more configuration for geo_client #357

• GEO: support GEOPOS API for redis proxy #369

• GEO: add GEOADD API for redis proxy to support full GEO function #381

Tools and Scripts:

• scripts: fix clusters_stat.sh and related scripts #375

• fix: improve shell get_app_stat and ls_nodes & fix scripts #382

Configuration updates:

[core]
+ tcmalloc_release_rate = 1.0 # default

[replication]
+ mem_release_interval_ms = 86400000 # default
+ mem_release_enabled = true  # default

[neverchanje]

Release Notes of v1.11.6

The following are the highlights in this release:

Important bugfix on failure detection!

This bug occurred a few times in a week in our production. Each time causes one server restart. For user who experiences this bug we suggest you to upgrade to 1.11.6.

Related PR: XiaoMi/rdsn#272, XiaoMi/rdsn#279, XiaoMi/rdsn#277

Related Issue: #315

Stability Fixes

Server connection threshold support: XiaoMi/rdsn#234

Whitelist support: XiaoMi/rdsn#228, XiaoMi/rdsn#226

Related Docs: https://pegasus-kv.github.io/administration/whitelist

Optimization of Memory Usage

In XiaoMi, some of our clusters experience unreasonably high memory load. To reduce the unknown-where memory cost we introduce "auto-memory-release", thanks to @linlinhaohao888.

Currently this feature is disable by default due to severe performance downgrade during each run of memory release.

Related PR: XiaoMi/rdsn#278

Related Docs: TBD

Upgrade from the previous version

[core]
- ;tool = fastrun
- cli_local = false
- cli_remote = false
- start_nfs = false
- ;logging_factory_name = dsn::tools::hpc_logger
+ tcmalloc_release_rate = 1.0 # default

- [tools.hpc_logger]
- per_thread_buffer_bytes = 8192

[threadpool.THREAD_POOL_DEFAULT]
- max_input_queue_length = 1024

[replication]
+ duplication_disabled = true
+ mem_release_interval_ms = 86400000 # 1day
+ mem_release_enabled = false # disable by default

[pegasus.server]
- updating_rocksdb_sstsize_interval_seconds = 600
+ update_rdb_stat_interval = 600

[task..default]
- fast_execution_in_network_thread = false