flink-extended/flink-remote-shuffle

Issues

• Support standby ShuffleManager

  #35 opened a year ago by TanYuxin-tyx
  0

• Implement ReducePartition

  #13 opened 3 years ago by wsry
  1

• SortBuffer supports reading data from the specified channel index

  #40 opened 2 years ago by TanYuxin-tyx
  0

• Conflict with Flink 1.15.4

  #94 opened 2 years ago by dangshazi
  2

• `SortBuffer` is blocked at `requestBufferFromPool`

  #93 opened 2 years ago by KevinLeeDan
  2

• flink 1.15.3 compile failed

  #92 opened 2 years ago by winnerhero
  0

• Further improvements umbrella for remote shuffle service

  #78 opened 2 years ago by TanYuxin-tyx
  0

• Adapt to Flink 1.16

  #83 opened 2 years ago by TanYuxin-tyx
  1

• Incompatibility with the new version of Flink

  #89 opened 2 years ago by TanYuxin-tyx
  0

• ShuffleReadClient is blocked when create channel

  #77 opened 2 years ago by dangshazi
  2

• flink 1.15.2 compile failed

  #87 opened 2 years ago by licc
  1

• Use apiextensions v1 instead of v1beta1 for remote shuffle operator crd

  #76 opened 2 years ago by wsry
  0

• Multi-tenant throttling and guarantee

  #86 opened 2 years ago by dangshazi
  0

• Add JVM CPU usage metirc

  #85 opened 2 years ago by dangshazi
  0

• Introduce new compression algorithms with higher compression ratio

  #81 opened 2 years ago by TanYuxin-tyx
  0

• Fix the bug that BaseMapPartition may not read data sequentially

  #79 opened 2 years ago by TanYuxin-tyx
  0

• Explicitly exit ShuffleWorker process when terminate future finished

  #75 opened 2 years ago by Aitozi
  1

• Fix the possible heartbeat issue caused by Zookeeper restart

  #72 opened 3 years ago by wsry
  0

• Support to config the shuffle cluster name prefix to use at Flink side

  #36 opened 3 years ago by wsry
  0

• Job can switch to new shuffle cluster only when the old cluster is unavailable

  #47 opened 3 years ago by wsry
  0

• Add more metrics

  #15 opened 3 years ago by wsry
  0

• Add more information about the ShuffleWorker in the ShuffleManagerClient

  #71 opened 3 years ago by Aitozi
  0

• Add the metrics reporter configuration in documents and modify the delimiter of the metric reporter classes

  #69 opened 3 years ago by TanYuxin-tyx
  0

• Support to restrict the maximum storage space can be used per disk

  #65 opened 3 years ago by wsry
  0

• Support to use StatefulSet + Pvc to deploy the shuffle service

  #66 opened 3 years ago by Aitozi
  2

• Metric ReadingThroughput missed

  #62 opened 3 years ago by dangshazi
  1

• SHUFFLE_CONF_DIR didn't work in standalone mode

  #59 opened 3 years ago by dangshazi
  2

• Avoid specifying "remote-shuffle.manager.rpc-address" when start ShuffleManager in standalone mode

  #60 opened 3 years ago by dangshazi
  1

• Upgrade log4j to the latest version

  #31 opened 3 years ago by wsry
  0

• Ensure all resources will be released for result partition and input gate when closed

  #55 opened 3 years ago by wsry
  0

• Not trigger fatal error if Flink job has finished when notifying new shuffle manager address

  #53 opened 3 years ago by wsry
  0

• Adapt to Flink 1.15

  #34 opened 3 years ago by wsry
  0

• Support Kubernetes HA Services

  #52 opened 3 years ago by czy006
  1

• Enable reading buffer floating for better fairness

  #50 opened 3 years ago by wsry
  0

• data request should implement crc check strategy

  #46 opened 3 years ago by lichaojacobs
  4

• Failed to start ShuffleWorker

  #42 opened 3 years ago by dangshazi
  3

• Introduce a simple disk health checker

  #28 opened 3 years ago by wsry
  1

• Optimize the round-robin load balance strategy and use it as the default

  #32 opened 3 years ago by wsry
  1

• Support to write shuffle data to local shuffle worker first

  #27 opened 3 years ago by wsry
  1

• Handle network issues gracefully

  #11 opened 3 years ago by wsry
  0

• Support data replication

  #29 opened 3 years ago by wsry
  0

• Load balance improvement

  #14 opened 3 years ago by wsry
  1

• Support dynamic log level

  #12 opened 3 years ago by wsry
  2

• Upgrade Log4j2 to 2.17.0

  #22 opened 3 years ago by wanglijie95
  6

• support pod templates in k8s deployment mode

  #19 opened 3 years ago by jlon
  1

• Provide more frequently used tools in container image

  #17 opened 3 years ago by wsry
  0

• The shuffle manager should restore the previously managed workers when re-election

  #8 opened 3 years ago by jlon
  6

• add PrometheusMetricReporterFactory

  #16 opened 3 years ago by jlon
  1

• Support Kubernetes-based high-availability services

  #9 opened 3 years ago by jlon
  3

• Provide a simple web UI

  #10 opened 3 years ago by wsry
  0