Draft
Conversation
…MemoryError in JDK21 This test started to fail quite regularly in JDK21. The problem was that the low heap size could have caused an OutOfMemoryError to appear when compiling the dummy classes. An OOM in the compilation phase results in a different error message being printed to stdout that wasn't captured by the test. The solution is to pre-compile the classes upfront (with the normal heap size). The test main method will only load the classes. No compilation is necessary.
…ated tests of DefaultSlotStatusSyncerTest Also deduplicate the code of these tests.
…ved before allocation success
(cherry picked from commit 95163e8)
…ob is suspended during Restarting phase
suspend and cancel reset the ExecutionGraph in a similar way. I move the common logic into its own method to make this more prominent in the code.
…out for AdaptiveSchedulerTest (apache#24400)
…Builder.cleanupInRocksdbCompactFilter
…3.11 wheel package This closes apache#24439.
…hod construct resource has exception. This closes apache#24462.
…ogs when the task exited. (apache#23922)
…used by multiple writes to the same sink table and shared staging directory This closes apache#24492 * Fix unstable TableSourceITCase#testTableHintWithLogicalTableScanReuse * Moves the staging dir configuration into builder for easier testing --------- Co-authored-by: Matthias Pohl <matthias.pohl@aiven.io> (cherry picked from commit 7d0111d)
…lerTest.testOnReadBufferRequestError
(cherry picked from commit 3342c23)
So far, we used a special value for the final checkpoint on endInput. However, as shown in the description of this ticket, final doesn't mean final. Hence, multiple committables with EOI could be created at different times. With this commit, we stop using a special value for such committables and instead try to guess the checkpoint id of the next checkpoint. There are various factors that influence the checkpoint id but we can mostly ignore them all because we just need to pick a checkpoint id that is - higher than all checkpoint ids of the previous, successful checkpoints of this attempt - higher than the checkpoint id of the restored checkpoint - lower than any future checkpoint id. Hence, we just remember the last observed checkpoint id (initialized with max(0, restored id)), and use last id + 1 for endInput. Naturally, multiple endInput calls happening through restarts will result in unique checkpoint ids. Note that aborted checkpoints before endInput may result in diverged checkpoint ids across subtasks. However, each of the id satisfies above requirements and any id of endInput1 will be smaller than any id of endInput2. Thus, diverged checkpoint ids will not impact correctness at all. (cherry picked from commit 9302545)
Co-authored-by: Ferenc Csaky <fcsaky@apache.org>
… literal without seconds (apache#26555)
…in adaptive scheduler Also enable this strategy by default via the introduced config option
…on via `CompiledPlan`
Co-authored-by: Matthias Pohl <github@mapohl.com>
…sabled for all connections unexpectedly
…Y type by evaluating Unsafe.arrayBaseOffset(byte[].class) in TM rather than in JM (apache#26592) Fix HashPartitioner codegen for BINARY/VARBINARY type by evaluating BYTE_ARRAY_BASE_OFFSET in TM instead of JM. The issue is, if JM memory is set > 32G while TM memory is set < 32G, this causes JVM to treat the JAVA process > 32G as large heap JVM. This can impact Unsafe behavior. For eg: UNSAFE.arrayBaseOffset(byte[].class) will return 24 for large heap JVM while 16 for others. Due to this, the tasks that run on TM (<32 G while JM > 32G or vice versa) that try to read the byte[] for MurmurHash read wrong memory locations. Signed-off-by: Jiangjie (Becket) Qin <becket.qin@gmail.com>
…es in batch mode (apache#27016) In apache#26433, we removed the EOI marker in the form of Long.MAX_VALUE as the checkpoint id. Since streaming pipelines can continue to checkpoint even after their respective operators have been shut down, it is not safe to use a constant as this can lead to duplicate commits. However, in batch pipelines we only have one commit on job shutdown. Using any checkpoint id should suffice in this scenario. Any pending committables should be processed by the ComitterOperator when the operator shuts down. No further checkpoints will take place. There are various connectors which rely on this behavior. I don't see any drawbacks from keeping this behavior for batch pipelines.
If a resource is lazily created in open, we can only close after checking for null. Otherwise a failure during initialization will trigger secondary failures.
…eckpoint notification is delayed (apache#27157)
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
20 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What is the purpose of the change
(For example: This pull request makes task deployment go through the blob server, rather than through RPC. That way we avoid re-transferring them on each deployment (during recovery).)
Brief change log
(for example:)
Verifying this change
Please make sure both new and modified tests in this PR follows the conventions defined in our code quality guide: https://flink.apache.org/contributing/code-style-and-quality-common.html#testing
(Please pick either of the following options)
This change is a trivial rework / code cleanup without any test coverage.
(or)
This change is already covered by existing tests, such as (please describe tests).
(or)
This change added tests and can be verified as follows:
(example:)
Does this pull request potentially affect one of the following parts:
@Public(Evolving): (yes / no)Documentation