[
https://issues.apache.org/jira/browse/FLINK-8360?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16365684#comment-16365684
]
ASF GitHub Bot commented on FLINK-8360:
---------------------------------------
Github user StefanRRichter commented on a diff in the pull request:
https://github.com/apache/flink/pull/5239#discussion_r168501090
--- Diff:
flink-runtime/src/main/scala/org/apache/flink/runtime/taskmanager/TaskManager.scala
---
@@ -1195,22 +1202,15 @@ class TaskManager(
config.getTimeout().getSize(),
config.getTimeout().getUnit()))
- //TODO: make configurable, this is just the future fallback case,
- // integrate with
ConfigConstants.TASK_MANAGER_LOCAL_STATE_ROOT_DIR_KEY
- // TODO: wire this so that the manager survives the end of the task
- //TODO create "localState" sub-directories!!!
- val taskExecutorLocalStateStoresManager =
- new
TaskExecutorLocalStateStoresManager(ioManager.getSpillingDirectories)
-
- val localStateStore =
taskExecutorLocalStateStoresManager.localStateStoreForTask(
+ val taskLocalStateStore =
taskManagerLocalStateStoresManager.localStateStoreForTask(
jobInformation.getJobId,
taskInformation.getJobVertexId,
tdd.getSubtaskIndex)
- val slotStateManager = new TaskStateManagerImpl(
+ val taskLocalStateManager = new TaskStateManagerImpl(
--- End diff --
👍
> Implement task-local state recovery
> -----------------------------------
>
> Key: FLINK-8360
> URL: https://issues.apache.org/jira/browse/FLINK-8360
> Project: Flink
> Issue Type: New Feature
> Components: State Backends, Checkpointing
> Reporter: Stefan Richter
> Assignee: Stefan Richter
> Priority: Major
> Fix For: 1.5.0
>
>
> This issue tracks the development of recovery from task-local state. The main
> idea is to have a secondary, local copy of the checkpointed state, while
> there is still a primary copy in DFS that we report to the checkpoint
> coordinator.
> Recovery can attempt to restore from the secondary local copy, if available,
> to save network bandwidth. This requires that the assignment from tasks to
> slots is as sticky is possible.
> For starters, we will implement this feature for all managed keyed states and
> can easily enhance it to all other state types (e.g. operator state) later.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
