* Parallel indexing single dim partitions
Implements single dimension range partitioning for native parallel batch
indexing as described in #8769. This initial version requires the
druid-datasketches extension to be loaded.
The algorithm has 5 phases that are orchestrated by the supervisor in
`ParallelIndexSupervisorTask#runRangePartitionMultiPhaseParallel()`.
These phases and the main classes involved are described below:
1) In parallel, determine the distribution of dimension values for each
input source split.
`PartialDimensionDistributionTask` uses `StringSketch` to generate
the approximate distribution of dimension values for each input
source split. If the rows are ungrouped,
`PartialDimensionDistributionTask.UngroupedRowDimensionValueFilter`
uses a Bloom filter to skip rows that would be grouped. The final
distribution is sent back to the supervisor via
`DimensionDistributionReport`.
2) The range partitions are determined.
In `ParallelIndexSupervisorTask#determineAllRangePartitions()`, the
supervisor uses `StringSketchMerger` to merge the individual
`StringSketch`es created in the preceding phase. The merged sketch is
then used to create the range partitions.
3) In parallel, generate partial range-partitioned segments.
`PartialRangeSegmentGenerateTask` uses the range partitions
determined in the preceding phase and
`RangePartitionCachingLocalSegmentAllocator` to generate
`SingleDimensionShardSpec`s. The partition information is sent back
to the supervisor via `GeneratedGenericPartitionsReport`.
4) The partial range segments are grouped.
In `ParallelIndexSupervisorTask#groupGenericPartitionLocationsPerPartition()`,
the supervisor creates the `PartialGenericSegmentMergeIOConfig`s
necessary for the next phase.
5) In parallel, merge partial range-partitioned segments.
`PartialGenericSegmentMergeTask` uses `GenericPartitionLocation` to
retrieve the partial range-partitioned segments generated earlier and
then merges and publishes them.
* Fix dependencies & forbidden apis
* Fixes for integration test
* Address review comments
* Fix docs, strict compile, sketch check, rollup check
* Fix first shard spec, partition serde, single subtask
* Fix first partition check in test
* Misc rewording/refactoring to address code review
* Fix doc link
* Split batch index integration test
* Do not run parallel-batch-index twice
* Adjust last partition
* Split ITParallelIndexTest to reduce runtime
* Rename test class
* Allow null values in range partitions
* Indicate which phase failed
* Improve asserts in tests
* add prefixes support to google input source, making it symmetrical-ish with s3
* docs
* more better, and tests
* unused
* formatting
* javadoc
* dependencies
* oops
* review comments
* better javadoc
* add s3 input source for native batch ingestion
* add docs
* fixes
* checkstyle
* lazy splits
* fixes and hella tests
* fix it
* re-use better iterator
* use key
* javadoc and checkstyle
* exception
* oops
* refactor to use S3Coords instead of URI
* remove unused code, add retrying stream to handle s3 stream
* remove unused parameter
* update to latest master
* use list of objects instead of object
* serde test
* refactor and such
* now with the ability to compile
* fix signature and javadocs
* fix conflicts yet again, fix S3 uri stuffs
* more tests, enforce uri for bucket
* javadoc
* oops
* abstract class instead of interface
* null or empty
* better error
* Fix the potential race SplittableInputSource.getNumSplits() and SplittableInputSource.createSplits() in TaskMonitor
* Fix docs and javadoc
* Add unit tests for large or small estimated num splits
* add override
* add parquet support to native batch
* cleanup
* implement toJson for sampler support
* better binaryAsString test
* docs
* i hate spellcheck
* refactor toMap conversion so can be shared through flattenerMaker, default impls should be good enough for orc+avro, fixup for merge with latest
* add comment, fix some stuff
* adjustments
* fix accident
* tweaks
If the JDBC drivers are missing from the lookup extensions, throw an
exception that directs the user how to resolve the issue. This change is
a follow up to #8825.
* transformSpec + array expressions
changes:
* added array expression support to transformSpec
* removed ParseSpec.verify since its only use afaict was preventing transform expr that did not replace their input from functioning
* hijacked index task test to test changes
* remove docs about being unsupported
* re-arrange test assert
* unused imports
* imports
* fix tests
* preserve types
* suppress warning, fixes, add test
* formatting
* cleanup
* better list to array type conversion and tests
* fix oops
* Add reference to `druid.storage.type`
This should be in here. Without setting storage type to S3 globally it will obviously not be used, even if all other parameters are correct.
* Update s3.md
Add global storage parameter to knob table.
* Update s3.md
* SQL: EARLIEST, LATEST aggregators.
I chose these names instead of FIRST, LAST because those are already
reserved functions in Calcite that mean something different. I think
these are also better names anyway.
* Finalify.
* SQL updates.
* Adjust aggregator calls.
* Validations, test updates.
* Review docs.
* sketch of broker parallel merges done in small batches on fork join pool
* fix non-terminating sequences, auto compute parallelism
* adjust benches
* adjust benchmarks
* now hella more faster, fixed dumb
* fix
* remove comments
* log.info for debug
* javadoc
* safer block for sequence to yielder conversion
* refactor LifecycleForkJoinPool into LifecycleForkJoinPoolProvider which wraps a ForkJoinPool
* smooth yield rate adjustment, more logs to help tune
* cleanup, less logs
* error handling, bug fixes, on by default, more parallel, more tests
* remove unused var
* comments
* timeboundary mergeFn
* simplify, more javadoc
* formatting
* pushdown config
* use nanos consistently, move logs back to debug level, bit more javadoc
* static terminal result batch
* javadoc for nullability of createMergeFn
* cleanup
* oops
* fix race, add docs
* spelling, remove todo, add unhandled exception log
* cleanup, revert unintended change
* another unintended change
* review stuff
* add ParallelMergeCombiningSequenceBenchmark, fixes
* hyper-threading is the enemy
* fix initial start delay, lol
* parallelism computer now balances partition sizes to partition counts using sqrt of sequence count instead of sequence count by 2
* fix those important style issues with the benchmarks code
* lazy sequence creation for benchmarks
* more benchmark comments
* stable sequence generation time
* update defaults to use 100ms target time, 4096 batch size, 16384 initial yield, also update user docs
* add jmh thread based benchmarks, cleanup some stuff
* oops
* style
* add spread to jmh thread benchmark start range, more comments to benchmarks parameters and purpose
* retool benchmark to allow modeling more typical heterogenous heavy workloads
* spelling
* fix
* refactor benchmarks
* formatting
* docs
* add maxThreadStartDelay parameter to threaded benchmark
* why does catch need to be on its own line but else doesnt
* Add option lateMessageRejectionStartDate
* Use option lateMessageRejectionStartDate
* Fix tests
* Add lateMessageRejectionStartDate to kafka indexing service
* Update tests kafka indexing service
* Fix tests for KafkaSupervisorTest
* Add lateMessageRejectionStartDate to KinesisSupervisorIOConfig
* Fix var name
* Update documentation
* Add check lateMessageRejectionStartDateTime and lateMessageRejectionPeriod, fails if both were specified.
* remove select query
* thanks teamcity
* oops
* oops
* add back a SelectQuery class that throws RuntimeExceptions linking to docs
* adjust text
* update docs per review
* deprecated
Since it hasn't received updates or community interest in a while, it makes sense
to de-emphasize it in the distribution and most documentation (outside of simple
mentions of its existence).
* Support assign tasks to run on different tiers of MiddleManagers
* address comments
* address comments
* rename tier to category and docs
* doc
* fix doc
* fix spelling errors
* docs
* Stateful auto compaction
* javaodc
* add removed test back
* fix test
* adding indexSpec to compactionState
* fix build
* add lastCompactionState
* address comments
* extract CompactionState
* fix doc
* fix build and test
* Add a task context to store compaction state; add javadoc
* fix it test
* IOConfig for compaction task
* add javadoc, doc, unit test
* fix webconsole test
* add spelling
* address comments
* fix build and test
* address comments
* Add tier based usage metrics for historical nodes to help with druid historical autoscaling
Add tier based usage metrics for historical nodes to help druid cluster orchestration systems understand the historical node usage and requirements. Following metrics would be helpful -
tier/required/capacity- total capacity in bytes required in each tier. Dimensions - tier
tier/total/capacity - total capacity in bytes available in a given tier. Dimension - tier
tier/historical/count - no. of historical nodes available in each tier. Dimension - tier
tier/replication/factor - configured maximum replication factor in given tier. Dimension - tier
* fix unit test failures
* Support LDAP authentication/authorization
* fixed integration-tests
* fixed Travis CI build errors related to druid-security module
* fixed failing test
* fixed failing test header
* added comments, force build
* fixes for strict compilation spotbugs checks
* removed authenticator rolling credential update feature
* removed escalator rolling credential update feature
* fixed teamcity inspection deprecated API usage error
* fixed checkstyle execution error, removed unused import
* removed cached config as part of removing authenticator rolling credential update feature
* removed config bundle entity as part of removing authenticator rolling credential update feature
* refactored ldao configuration
* added support for SSLContext configuration and TLSCertificateChecker
* removed check to return authentication failure when user has no group assigned, will be checked and handled by the authorizer
* Separate out authorizer checks between metadata-backed store user and LDAP user/groups
* refactored BasicSecuritySSLSocketFactory usage to fix strict compilation spotbugs checks
* fixes build issue
* final review comments updates
* final review comments updates
* fixed LGTM and spellcheck alerts
* Fixed Avatica auth failure error message check
* Updated metadata credentials validator exception message string, replaced DB with metadata store
* groupBy query: optional limit push down to segment scan
* make segment level limit push down configurable
* fix teamcity errors
* fix segment limit pushdown flag handling on query level config override
* use equals for comparator check
* fix sql and null handling
* fix unused imports
* handle null offset in NullableValueGroupByColumnSelectorStrategy for buffer comparator similar to RowBasedGrouperHelper.NullableRowBasedKeySerdeHelper
* fix segment underReplicated/unavailable counts to be gauges instead of counters
* fix jvm/gc/cpu to be a counter instead of timre
jvm/gc/cpu represents the total cpu time spent for multiple gc
invocations, not the time spent in each gc cycle.
the number needs to be divided by jvm/gc/count to get the average gc
time per cycle
* update docs
* fix spellcheck
* Implementing dropwizard emitter for druid
making metric manager and alert emitters as optional
* Refactor and make things work
more improvements
improve docs
refactrings
* Fix teamcity inspections
* review comments
* more review comments
* add limit to max number of gauges
* update pom version
* fix pom
* review comments
* review comment
* review comments
* fix broken doc link
review comments
review comments
* review comments
* fix checkstyle
* more spell check fixes
* fix travis failures
* #7641 - Changing segment distribution algorithm to distribute segments to multiple segment cache locations
* Fixing indentation
* WIP
* Adding interface for location strategy selection, least bytes used strategy impl, round-robin strategy impl, locationSelectorStrategy config with least bytes used strategy as the default strategy
* fixing code style
* Fixing test
* Adding a method visible only for testing, fixing tests
* 1. Changing the method contract to return an iterator of locations instead of a single best location. 2. Check style fixes
* fixing the conditional statement
* Added testSegmentDistributionUsingLeastBytesUsedStrategy, fixed testSegmentDistributionUsingRoundRobinStrategy
* to trigger CI build
* Add documentation for the selection strategy configuration
* to re trigger CI build
* updated docs as per review comments, made LeastBytesUsedStorageLocationSelectorStrategy.getLocations a synchronzied method, other minor fixes
* In checkLocationConfigForNull method, using getLocations() to check for null instead of directly referring to the locations variable so that tests overriding getLocations() method do not fail
* Implementing review comments. Added tests for StorageLocationSelectorStrategy
* Checkstyle fixes
* Adding java doc comments for StorageLocationSelectorStrategy interface
* checkstyle
* empty commit to retrigger build
* Empty commit
* Adding suppressions for words leastBytesUsed and roundRobin of ../docs/configuration/index.md file
* Impl review comments including updating docs as suggested
* Removing checkLocationConfigForNull(), @NotEmpty annotation serves the purpose
* Round robin iterator to keep track of the no. of iterations, impl review comments, added tests for round robin strategy
* Fixing the round robin iterator
* Removed numLocationsToTry, updated java docs
* changing property attribute value from tier to type
* Fixing assert messages
* Added live reports for Kafka and Native batch task
* Removed unused local variables
* Added the missing unit test
* Refine unit test logic, add implementation for HttpRemoteTaskRunner
* checksytle fixes
* Update doc descriptions for updated API
* remove unnecessary files
* Fix spellcheck complaints
* More details for api descriptions
* Adjust defaults for hashed partitioning
If neither the partition size nor the number of shards are specified,
default to partitions of 5,000,000 rows (similar to the behavior of
dynamic partitions). Previously, both could be null and cause incorrect
behavior.
Specifying both a partition size and a number of shards now results in
an error instead of ignoring the partition size in favor of using the
number of shards. This is a behavior change that makes it more apparent
to the user that only one of the two properties will be honored
(previously, a message was just logged when the specified partition size
was ignored).
* Fix test
* Handle -1 as null
* Add -1 as null tests for single dim partitioning
* Simplify logic to handle -1 as null
* Address review comments
* Rename partition spec fields
Rename partition spec fields to be consistent across the various types
(hashed, single_dim, dynamic). Specifically, use targetNumRowsPerSegment
and maxRowsPerSegment in favor of targetPartitionSize and
maxSegmentSize. Consistent and clearer names are easier for users to
understand and use.
Also fix various IntelliJ inspection warnings and doc spelling mistakes.
* Fix test
* Improve docs
* Add targetRowsPerSegment to HashedPartitionsSpec
* move google ext docs from contrib to core
* fix links
* revert unintended change
* more links, add note to example ext doc that it was removed, unlink from sidebar
* Exit JVM on curator unhandled errors
If an unhandled error occurs when curator is talking to ZooKeeper, exit
the JVM in addition to stopping the lifecycle to prevent the process
from being left in a zombie state. With this change,
BoundedExponentialBackoffRetryWithQuit is no longer needed as when
curator exceeds the configured retries, it triggers its unhandled error
listeners. A new "connectionTimeoutMs" CuratorConfig setting is added
mostly to facilitate testing curator unhandled errors, but it may be
useful for users as well.
* Address review comments
* Add realization for updating version of derived segments in MaterializedView
* add unit test, and change code style for the sake of ease of understanding
* fix document's mistake of expression
* Zookeeper version is updated.
* Zookeeper version is updated at licenses.yaml
* licenses.yaml is updated and dependencies are fixed to make the project successfully build.
* Zookeeper versions are fixed at licenses.yaml
* Add group_id to overlord tasks API and sys.tasks table
* adjust test
* modify docs
* Make groupId nullable
* fix integration test
* fix toString
* Remove groupId from TaskInfo
* Modify docs and tests
* modify TaskMonitorTest
* migrate binary notice entries to live in licenses.yaml, use licenses.yaml and NOTICE to generate NOTICE.BINARY at distribution time
* +x
* move release scripts to distribution/bin, fixup notice script, trim dependencies for avro and kerberos in licenses.yaml
* add missing hdfs-storage dependencies
* revert to old syntax, fixes
* formatting
* update notices for recently updated dependencies
Himanshu
Benjamin Hopp
Jihoon Son
Parag Jain
Chi Cao Minh
Vadim Ogievetsky
Roman Leventov
Clint Wylie
Jonathan Wei
Lucas Capistrant
Fangyuan Deng
SeKing
Surekha
fst0
Gian Merlino
Jad Naous
Tijo Thomas
Giuseppe Martino
Eyal Yurman
David Glasser
Abhishek Radhakrishnan
Kamal Gurala
Mingming Qiu
Mitch Lloyd
Nishant Bangarwa
Mohammad J. Khan
Xavier Léauté
pdeva
Sashidhar Thallam
Rye
legendtkl
Furkan KAMACI
Fokko Driesprong