[
https://issues.apache.org/jira/browse/HUDI-5408?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
sivabalan narayanan closed HUDI-5408.
-------------------------------------
Resolution: Fixed
> Partially failed commits in MDT have to be rolled back in all cases
> -------------------------------------------------------------------
>
> Key: HUDI-5408
> URL: https://issues.apache.org/jira/browse/HUDI-5408
> Project: Apache Hudi
> Issue Type: Bug
> Components: metadata
> Reporter: sivabalan narayanan
> Assignee: sivabalan narayanan
> Priority: Blocker
> Labels: pull-request-available
> Fix For: 0.13.0
>
>
> when compaction failed after completing in MDT but before completing in DT.
> and later when we re-attempt to apply the same compaction instant to MDT, we
> might miss to rollback any partially failed commit in MDT.
> Code of interest in SparkHoodieBackedTableMetadataWriter:
> {code:java}
> if (!metadataMetaClient.getActiveTimeline().containsInstant(instantTime)) {
> // if this is a new commit being applied to metadata for the first time
> writeClient.startCommitWithTime(instantTime);
> } else {
> Option<HoodieInstant> alreadyCompletedInstant =
> metadataMetaClient.getActiveTimeline().filterCompletedInstants().filter(entry
> -> entry.getTimestamp().equals(instantTime)).lastInstant();
> if (alreadyCompletedInstant.isPresent()) {
> // this code path refers to a re-attempted commit that got committed to
> metadata table, but failed in datatable.
> // for eg, lets say compaction c1 on 1st attempt succeeded in metadata
> table and failed before committing to datatable.
> // when retried again, data table will first rollback pending compaction.
> these will be applied to metadata table, but all changes
> // are upserts to metadata table and so only a new delta commit will be
> created.
> // once rollback is complete, compaction will be retried again, which
> will eventually hit this code block where the respective commit is
> // already part of completed commit. So, we have to manually remove the
> completed instant and proceed.
> // and it is for the same reason we enabled
> withAllowMultiWriteOnSameInstant for metadata table.
> HoodieActiveTimeline.deleteInstantFile(metadataMetaClient.getFs(),
> metadataMetaClient.getMetaPath(), alreadyCompletedInstant.get());
> metadataMetaClient.reloadActiveTimeline();
> }
> // If the alreadyCompletedInstant is empty, that means there is a requested
> or inflight
> // instant with the same instant time. This happens for data table clean
> action which
> // reuses the same instant time without rollback first. It is a no-op here
> as the
> // clean plan is the same, so we don't need to delete the requested and
> inflight instant
> // files in the active timeline.
> } {code}
> incase of else block, if there happen to be a partially failed commit in MDT,
> we may miss to roll it back.
> we might need to fix the flow.
>
> Imp to consider:
> even before attempting compaction, we should ensure there are no partially
> failed commits in MDT. If not, we need to ensure we consider list of valid
> instants while executing the compaction.
>
> Impact:
> some invalid data blocks will be considered valid since we fail to do eager
> rollbacks.
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
