Releases: Tencent/Firestorm
Releases · Tencent/Firestorm
release-0.4.1
[Bugfix] Fix quorum bug in blockIds report (#107)
release-0.4.0
What's Changed
- [Bugfix] Fix uncorrect index file by @jerqi in #92
- [Doc] Add the support information by @jerqi in #95
- [Bugfix] Exist memory leak when we use aqe by @jerqi in #98
- [Bugfix] Add timestamp to avoid the conflicts of gc.log after server … by @frankliee in #99
- [Improvement] Remove some integration tests under StorageType.HDFS mode to … by @frankliee in #97
- [Feature] Firestorm supports quorum write/read by @frankliee in #96
- upgrade to 0.4.0 by @jerqi in #102
Full Changelog: release-0.3.0...release-0.4.0
release-0.3.0
What's Changed
- [Feature] Add AccessManager by @duanmeng in #61
- [Bug] Exchange the position between
startHeartbeat
andregisterShuffleServers
by @jerqi in #75 - [Improvement] Add detailed logging for read client by @frankliee in #72
- [Feature]Access check in client by @duanmeng in #66
- [Feature] Add the metrics of requireBuffer failures by @frankliee in #78
- [Feature] Enhance the DelegationRssShuffleManager to support hdfs conf file and reset the shuffle manager name by @duanmeng in #79
- [Improvement] Update spark.rss.client.read.buffer.size to avoid JVM humongous alloc… by @frankliee in #80
- [Improvement] Update some default values of parameters by @frankliee in #82
- [Feature] add the dynamic client conf fetching and updating by @duanmeng in #81
- [Feature] add access related metrics by @duanmeng in #84
- [Feature] Support Spark 3.2 by @jerqi in #86
- [Feature] corrupted LocalStorage can be excluded by @jerqi in #85
- upgrade to 0.3.0 by @duanmeng in #89
Full Changelog: release-0.2.0...release-0.3.0
release-0.2.0
What's Changed
- Add license header to maven ci yaml file #2 by @duanmeng in #2
- Add pull request template by @duanmeng in #10
- Upgrade version to 0.2.0-SNAPSHOT by @duanmeng in #11
- [Improvement] Replace runtime exception with rss related exception in… by @colinmjj in #8
- Fix Dependabot alerts by @duanmeng in #12
- [Feature] Read index file first in local mode by @duanmeng in #6
- [Feature] Add modifier, comment and import checking into checkstyle by @duanmeng in #16
- [Feature] Enable NewLineAtEofChecker in checkstyle by @duanmeng in #18
- [Feature] Firestorm supports node's health check by @jerqi in #14
- [Feature] Add the metrics for the number of total partitions by @jerqi in #24
- [Feature] Limit upload segment num and upload time in force mode by @duanmeng in #23
- Add metrics to collect connection info of grpc by @colinmjj in #25
- [Improvement] Fix the flaky test by @jerqi in #29
- [Improvement] Support to config byte sizes by readable string (e.g., kb/mb/gb) by @frankliee in #26
- [Feature] Add doc system based on Jekyll by @frankliee in #22
- [Feature] Read index file first in hdfs mode by @duanmeng in #21
- [Improvement] Modify some configuration's names and default values by @jerqi in #27
- Fix log4j-core dependabot by @duanmeng in #32
- [Improvement]Remove useless code in test base by @duanmeng in #33
- [Improvement] Avoid speculative tasks causing writing errors after shuffle is uploaded and deleted by @jerqi in #31
- [Improvement] fix log4j Dependabot alerts by @duanmeng in #37
- [Feature]Read handler refactor by @duanmeng in #34
- [MINOR] Fix flaky testes in HealthCheckCoordinatorGrpcTest by @frankliee in #38
- Check blockIds in read client to allow early return by @frankliee in #39
- Bump log4j-core from 2.16.0 to 2.17.0 by @dependabot in #44
- [Improvement] Support Firestorm with Spark 2.3 & 3.0 by @frankliee in #41
- Support store shuffle data in memory by @colinmjj in #36
- Update Readme with memory shuffle support by @colinmjj in #46
- [Improvement] Update import style in the Spark-like way by @frankliee in #45
- Shutdown non-daemon executor service to avoid hang when jvm destroy by @toujours33 in #47
- Fix time unit from micro to mill by @colinmjj in #52
- upload unit test log in github CI by @frankliee in #54
- [Improvement] Remove unnecssary logic to calculate bitmap split number by @colinmjj in #53
- [Feature] MultiStorageManager can write both localfile and hdfs directly by @jerqi in #43
- [Build] Only upload logs when test cases are failed by @frankliee in #57
- [Bug]Fix memory leak in spark executor by @colinmjj in #58
- [Feature] Support to skip unexpected and processed segments in LocalFileReadClient by @frankliee in #55
- [Feature] Support read from localfile, hdfs and uploaded hdfs by @jerqi in #56
- Fix log4j-core dependabot, upgrade to 2.17.1 by @frankliee in #62
- [Improvement] Reduce memory cost in spark executor by @colinmjj in #59
- [Feature] Hdfs write fail, the data will write to localfile by @jerqi in #60
- [Feature] Support StorageType MEMORY_LOCALFILE_HDFS by @jerqi in #63
- [Bug] Support to read incomplete index data by @jerqi in #67
- [Bug] LocalStorageChecker should support MEMORY_LOCALFILE and MEMORY_LOCALFILE_HDFS by @jerqi in #65
- [Feature] Add the data size of metrics HDFS and LOCALFILE by @jerqi in #68
- [Follow Up] Support to skip unexpected and processed segments in HDFS by @frankliee in #64
- [Bug] Replace parallelStream with stream except for the method
sendDataAsync
by @jerqi in #69 - Update version to 0.2.0 by @colinmjj in #74
- [Doc] Update readme for new configuration by @colinmjj in #73
New Contributors
- @duanmeng made their first contribution in #2
- @colinmjj made their first contribution in #8
- @frankliee made their first contribution in #26
- @dependabot made their first contribution in #44
- @toujours33 made their first contribution in #47
Full Changelog: https://github.com/Tencent/Firestorm/commits/release-0.2.0
release-0.1.0
Upgrade the version to 0.1.0