The previously used GCS API client library returned last update time for objects directly in milliseconds. The new library returns it in OffsetDateTime format which was being converted to seconds and stored against the object. This fix converts the time back to ms before storing it.
Changes:
- Add visibility into number of segments read/published by each parallel compaction
- Add new fields `segmentsRead`, `segmentsPublished` to `IngestionStatsAndErrorsTaskReportData`
- Update `ParallelIndexSupervisorTask` to populate the new stats
* initial commit
* comments
* typo
* comments
* comments
* remove var
* initialize global var early
* remove new line
* small test fix
* same fix another test
BaseNodeRoleWatcher counts down cacheInitialized after a timeout, but also sets some flag that it was a timed-out initialization. and call nodeViewInitializationTimedOut (new method on listeners) instead of nodeViewInitialized. Then listeners can do what is most appropriate with this information.
* Move retries into DataSegmentPusher implementations.
The individual implementations know better when they should and should
not retry. They can also generate better error messages.
The inspiration for this patch was a situation where EntityTooLarge was
generated by the S3DataSegmentPusher, and retried uselessly by the
retry harness in PartialSegmentMergeTask.
* Fix missing var.
* Adjust imports.
* Tests, comments, style.
* Remove unused import.
Changes:
- Add visibility into number of records processed by each streaming task per partition
- Add field `recordsProcessed` to `IngestionStatsAndErrorsTaskReportData`
- Populate number of records processed per partition in `SeekableStreamIndexTaskRunner`
* Globally disable AUTO_CLOSE_JSON_CONTENT.
This JsonGenerator feature is on by default. It causes problems with code
like this:
try (JsonGenerator jg = ...) {
jg.writeStartArray();
for (x : xs) {
jg.writeObject(x);
}
jg.writeEndArray();
}
If a jg.writeObject call fails due to some problem with the data it's
reading, the JsonGenerator will write the end array marker automatically
when closed as part of the try-with-resources. If the generator is writing
to a stream where the reader does not have some other mechanism to realize
that an exception was thrown, this leads the reader to believe that the
array is complete when it actually isn't.
Prior to this patch, we disabled AUTO_CLOSE_JSON_CONTENT for JSON-wrapped
SQL result formats in #11685, which fixed an issue where such results
could be erroneously interpreted as complete. This patch fixes a similar
issue with task reports, and all similar issues that may exist elsewhere,
by disabling the feature globally.
* Update test.
During ingestion, incremental segments are created in memory for the different time chunks and persisted to disk when certain thresholds are reached (max number of rows, max memory, incremental persist period etc). In the case where there are a lot of dimension and metrics (1000+) it was observed that the creation/serialization of incremental segment file format for persistence and persisting the file took a while and it was blocking ingestion of new data. This affected the real-time ingestion. This serialization and persistence can be parallelized across the different time chunks. This update aims to do that.
The patch adds a simple configuration parameter to the ingestion tuning configuration to specify number of persistence threads. The default value is 1 if it not specified which makes it the same as it is today.
* Merge hydrant runners flatly for realtime queries.
Prior to this patch, we have two layers of mergeRunners for realtime
queries: one for each Sink (a logical segment) and one across all
Sinks. This is done because to keep metrics and results grouped by Sink,
given that each FireHydrant within a Sink has its own separate storage
adapter.
However, it costs extra memory usage due to the extra layer of
materialization. This is especially pronounced for groupBy queries,
which only use their merge buffers at the top layer of merging. The
lower layer of merging materializes ResultRows directly into the heap,
which can cause heap exhaustion if there are enough ResultRows.
This patch changes to a single layer of merging when bySegment: false,
just like Historicals. To accommodate that, segment metrics like
query/segment/time are now per-FireHydrant instead of per-Sink.
Two layers of merging are retained when bySegment: true. This isn't
common, because it's typically only used when segment level caching
is enabled on the Broker, which is off by default.
* Use SinkQueryRunners.
* Remove unused method.
* Kill tasks should honor the buffer period of unused segments.
- The coordinator duty KillUnusedSegments determines an umbrella interval
for each datasource to determine the kill interval. There can be multiple unused
segments in an umbrella interval with different used_status_last_updated timestamps.
For example, consider an unused segment that is 30 days old and one that is 1 hour old. Currently
the kill task after the 30-day mark would kill both the unused segments and not retain the 1-hour
old one.
- However, when a kill task is instantiated with this umbrella interval, it’d kill
all the unused segments regardless of the last updated timestamp. We need kill
tasks and RetrieveUnusedSegmentsAction to honor the bufferPeriod to avoid killing
unused segments in the kill interval prematurely.
* Clarify default behavior in docs.
* test comments
* fix canDutyRun()
* small updates.
* checkstyle
* forbidden api fix
* doc fix, unused import, codeql scan error, and cleanup logs.
* Address review comments
* Rename maxUsedFlagLastUpdatedTime to maxUsedStatusLastUpdatedTime
This is consistent with the column name `used_status_last_updated`.
* Apply suggestions from code review
Co-authored-by: Kashif Faraz <kashif.faraz@gmail.com>
* Make period Duration type
* Remove older variants of runKilLTask() in OverlordClient interface
* Test can now run without waiting for canDutyRun().
* Remove previous variants of retrieveUnusedSegments from internal metadata storage coordinator interface.
Removes the following interface methods in favor of a new method added:
- retrieveUnusedSegmentsForInterval(String, Interval)
- retrieveUnusedSegmentsForInterval(String, Interval, Integer)
* Chain stream operations
* cleanup
* Pass in the lastUpdatedTime to markUnused test function and remove sleep.
---------
Co-authored-by: Kashif Faraz <kashif.faraz@gmail.com>
* Clear "lineSplittable" for JSON when using KafkaInputFormat.
JsonInputFormat has a "withLineSplittable" method that can be used to
control whether JSON is read line-by-line, or as a whole. The intent
is that in streaming ingestion, "lineSplittable" is false (although it
can be overridden by "assumeNewlineDelimited"), and in batch ingestion,
lineSplittable is true.
When a "json" format is wrapped by a "kafka" format, this isn't set
properly. This patch updates KafkaInputFormat to set this on an
underlying "json" format.
The tests for KafkaInputFormat were overriding the "lineSplittable"
parameter explicitly, which wasn't really fair, because that made them
unrealistic to what happens in production. Now they omit the parameter
and get the production behavior.
* Add test.
* Fix test coverage.
* Faster parsing: reduce String usage, list-based input rows.
Three changes:
1) Reworked FastLineIterator to optionally avoid generating Strings
entirely, and reduce copying somewhat. Benefits the line-oriented
JSON, CSV, delimited (TSV), and regex formats.
2) In the delimited (TSV) format, when the delimiter is a single byte,
split on UTF-8 bytes directly.
3) In CSV and delimited (TSV) formats, use list-based input rows when
the column list is provided upfront by the user.
* Fix style.
* Fix inspections.
* Restore validation.
* Remove fastutil-extra.
* Exception type.
* Fixes for error messages.
* Fixes for null handling.
Changes:
- Add new task context flag useConcurrentLocks.
- This can be set for an individual task or at a cluster level using `druid.indexer.task.default.context`.
- When set to true, any appending task would use an APPEND lock and any other
ingestion task would use a REPLACE lock when using time chunk locking.
- If false (default), we fall back on the context flag taskLockType and then useSharedLock.
Changes:
- Handle exception in deletePendingSegments API and map to correct HTTP status code
- Clean up exception message using `DruidException`
- Add unit tests
Currently, If 2 tasks are consuming from the same partitions, try to publish the segment and update the metadata, the second task can fail because the end offset stored in the metadata store doesn't match with the start offset of the second task. We can fix this by retrying instead of failing.
AFAIK apart from the above issue, the metadata mismatch can happen in 2 scenarios:
- when we update the input topic name for the data source
- when we run 2 replicas of ingestion tasks(1 replica will publish and 1 will fail as the first replica has already updated the metadata).
Implemented the comparable function to compare the last committed end offset and new Sequence start offset. And return a specific error msg for this.
Add retry logic on indexers to retry for this specific error msg.
Updated the existing test case.
The initial step in optimizing segment metadata was to centralize the construction of datasource schema in the Coordinator (#14985). Subsequently, our goal is to eliminate the requirement for regularly executing queries to obtain segment schema information. This task encompasses addressing both realtime and finalized segments.
This modification specifically addresses the issue with realtime segments. Tasks will now routinely communicate the schema for realtime segments during the segment announcement process. The Coordinator will identify the schema alongside the segment announcement and subsequently update the schema for realtime segments in the metadata cache.
Changes
- Audit create or update of a supervisor spec. The purpose of the audit is
to track which user made change to a supervisor and when.
- The audit entry does not contain the entire spec or even a diff of the changes
as this is already captured in the `druid_supervisors` metadata table.
Description
When batchAllocationWaitTime is set to 0, the segment allocation queue is polled continuously even when it is empty. This would take up cpu cycles unnecessarily.
Some existing race conditions would also become more frequent when the batchAllocationWaitTime is 0. This PR tries to better address those race conditions as well.
Changes
Do not reschedule a poll if queue is empty
When a new batch is added to queue, schedule a poll
Simplify keyToBatch map
Handle race conditions better
As soon as a batch starts getting processed, do not add any more requests to it
* Allow empty inserts and replace.
- Introduce a new query context failOnEmptyInsert which defaults to false.
- When this context is false (default), MSQE will now allow empty inserts and replaces.
- When this context is true, MSQE will throw the existing InsertCannotBeEmpty MSQ fault.
- For REPLACE ALL over an ALL grain segment, the query will generate a tombstone spanning eternity
which will be removed eventually be the coordinator.
- Add unit tests in MSQInsertTest, MSQReplaceTest to test the new default behavior (i.e., when failOnEmptyInsert = false)
- Update unit tests in MSQFaultsTest to test the non-default behavior (i.e., when failOnEmptyInsert = true)
* Ignore test to see if it's the culprit for OOM
* Add heap dump config
* Bump up -Xmx from 1500 MB to 2048 MB
* Add steps to tarball and collect hprof dump to GHA action
* put back mx to 1500MB to trigger the failure
* add the step to reusable unit test workflow as well
* Revert the temp heap dump & @Ignore changes since max heap size is increased
* Minor updates
* Review comments
1. Doc suggestions
2. Add tests for empty insert and replace queries with ALL grain and limit in the
default failOnEmptyInsert mode (=false). Add similar tests to MSQFaultsTest with
failOnEmptyInsert = true, so the query does fail with an InsertCannotBeEmpty fault.
3. Nullable annotation and javadocs
* Add comment
replace_limit.patch
Currently in the realtime ingestion (Kafka/Kinesis) case, after publishing the segments, upon acknowledgement from the coordinator that the segments are already placed in some historicals, the peon would unannounce the segments (basically saying the segments are not in this peon anymore to the whole cluster) and drop the segments from cache and sink timeline in one shot.
The in transit queries from the brokers that still thinks the segments are in the peon can get a NullPointer exception when the peon is unsetting the hydrants in the sinks.
The fix would let the peon to wait for a configurable delay period before dropping segments, remove segments from cache etc after the peon unannounce the segments.
This delayed approach is similar to how the historicals handle segments moving out.
Changes
- Add `log` implementation for `AuditManager` alongwith `SQLAuditManager`
- `LoggingAuditManager` simply logs the audit event. Thus, it returns empty for
all `fetchAuditHistory` calls.
- Add new config `druid.audit.manager.type` which can take values `log`, `sql` (default)
- Add new config `druid.audit.manager.logLevel` which can take values `DEBUG`, `INFO`, `WARN`.
This gets activated only if `type` is `log`.
- Remove usage of `ConfigSerde` from `AuditManager` as audit is not just limited to configs
- Add `AuditSerdeHelper` for a single implementation of serialization/deserialization of
audit payload and other utility methods.
The segment allocation algorithm reuses an already allocated pending segment if the new allocation request is made for the same parameters:
datasource
sequence name
same interval
same value of skipSegmentLineageCheck (false for batch append, true for streaming append)
same previous segment id (used only when skipSegmentLineageCheck = false)
The above parameters can thus uniquely identify a pending segment (enforced by the UNIQUE constraint on the sequence_name_prev_id_sha1 column in druid_pendingSegments metadata table).
This reuse is done in order to
allow replica tasks (in case of streaming ingestion) to use the same set of segment IDs.
allow re-run of a failed batch task to use the same segment ID and prevent unnecessary allocations
* Fixing failing compaction/parallel index jobs during upgrade due to new actions not available on the overlord.
* Fixing build
* Removing extra space.
* Fixing json getter.
* Review comments.
Changes:
- Fix log `Got end of partition marker for partition [%s] from task [%s] in discoverTasks`
by fixing order of args
- Simplify in-line classes by using lambda
- Update kill task message from `Task [%s] failed to respond to [set end offsets]
in a timely manner, killing task` to `Failed to set end offsets, killing task`
- Clean up tests
Fixes missing task failure error message on Overlord.
The error message was missing since TaskManagementResource#assignTask API wasn't annotated with @Produces(MediaType.APPLICATION_JSON) resulting in the response being treated as application/octet-stream, that in turn lead to MessageBodyWriter not found error on the middle manager. The exception is not logged on the middle manager itself since it happens even before entering the assignTask function -- while mapping arg Task -> MSQControllerTask.
Currently when we submit a task to druid and number of currently active tasks has already reached (druid.indexer.queue.maxSize) then 500 ISE is thrown as per shown in the screenshot in #15380.
This fix will return HTTP 429 Too Many Requests(with proper error message) instead of 500 ISE, when we submit a task and queueSize has reached.
The TaskQueue maintains a map of active task ids to tasks, which can be utilized to get active task payloads, before falling back to the metadata store.
* MSQ generates tombstones honoring the query's granularity.
This change tweaks to only account for the infinite-interval tombstones.
For finite-interval tombstones, the MSQ query granualrity will be used
which is consistent with how MSQ works.
* more tests and some cleanup.
* checkstyle
* comment edits
* Throw TooManyBuckets fault based on review; add more tests.
* Add javadocs for both methods on reconciling the methods.
* review: Move testReplaceTombstonesWithTooManyBucketsThrowsException to MsqFaultsTest
* remove unused imports.
* Move TooManyBucketsException to indexing package for shared exception handling.
* lower max bucket for tests and fixup count
* Advance and count the iterator.
* checkstyle
Vishesh Garg
George Shiqi Wu
Adithya Chakilam
AmatyaAvadhanula
Gian Merlino
Sensor
Zoltan Haindrich
YongGang
Adarsh Sanjeev
Pramod Immaneni
Sam Wheating
Karan Kumar
Abhishek Radhakrishnan
Kashif Faraz
PANKAJ KUMAR
Rishabh Singh
kaisun2000
TestBoost
Clint Wylie
Sachidananda Maharana
Krishna Anandan