beam icon indicating copy to clipboard operation
beam copied to clipboard
verified publisher icon apache

[Tracking Umbrella] Prism Runner areas for contribution.

Open lostluck opened this issue 2 years ago • 0 comments

What needs to happen?

This issue is to track and refer to other issues/prs for various prism features. This issue shouldn't generally be commented on, but have this top entry edited as needed, referring to granular issues for individual features and support.

Ultimately, this will eventually track support in the Beam Compatibility Matrix, and keeping the Prism README up to date.

Complete items should be checked, and have links to their completing PR or closed primary tracking issue.

Items marked should only have an issue filed when the work has started, typically there's a meaningful design proposal, and understanding of what the closing criteria are. This can be "X set of existing SDK tests now pass", or a given capability is possible (eg. UI related features.)

Prism Areas for Contribution

Beam Core Priorities

These are features that prevent Prism use and adoption.

In progress by @lostluck

  • [x] State Handling (#28543)
  • [x] #29772
    • [x] Event Time
    • [x] #30083
      • [ ] #32786
    • [x] #32115
    • [x] #32121
    • [ ] #32211
  • [x] #29917
    • [x] Watermark and Element Event handling
    • [x] #30083
  • [ ] #31438
    • [x] #31463
    • [x] #31462
    • [ ] #31461

Beam Feature Burn Down (from Java and Python Validates Runner Tests)

The goal in this section is correctly implement Beam features in prism such that the validates runner suites of each SDK pass. The following issues were produced from examining the test outputs from the failing tests. This list will be refined as new failures are discovered initial blocking features are implemented.

  • Features

    • [x] #31912
    • [ ] #31438
      • [ ] [prism] Support WindowingStrategy.AccumulationMode Accumulating/Discarding
      • [ ] [prism] Support WindowingStrategy.AllowedLateness
    • [x] #31462
    • [ ] #31921
    • [x] #32115

  • Metrics

    • [x] #31926
      • MetricsTest$AttemptedMetricTests.testAttemptedDistributionMetrics
      • MetricsTest$AttemptedMetricTests.testAttemptedCounterMetrics
    • [x] #31927
      • MetricsTest.testCommittedStringSetMetrics
      • MetricsTest.testAttemptedStringSetMetrics
    • [x] #31928
      • MetricsTest.testCommittedGuageMetrics
      • MetricsTest.testAttemptedGuageMetrics

  • Windowing

    • [x] #32085

  • Various Coders or PreProcessing issues.

    • [x] #31991
    • [x] #32003
    • [x] #31992
    • [x] #32004
    • [x] [prism] Java CoGgroupByKeyTest - test failures (likely coders)
    • [ ] [prism] Java FlattenTest.testFlattenMultipleCoders - worker crash
    • [ ] [prism] Java org.apache.beam.sdk.transforms.GroupByKeyTest$BasicTests - testLargeKeys100MB
      • Now a mystery panic? error? Message is getting lost. Need to find what's getting logged.
    • [ ] ViewTest.testSideInputWithNestedIterables - user coder error
    • [ ] org.apache.beam.sdk.transforms.ParDoSchemaTest - UserCoder not a row coder exceptions.
      • Likely started failing due to other fixes. Might need to change Row Coder handling in some instances, instead of simply wrapping them.
    • [ ] #32636

  • State and Timer issues.

    • [x] #32064
    • [x] [prism] Python Assert failures - Missing Data (likely state & timers)
    • [ ] #32211
    • [x] #32221
    • [ ] [prism] Support SortedTime input (java annotation & requirement)
    • [ ] [prism] Support OrderedListState
    • [ ] #32222

  • Next Steps - only once the filtered suite fully passes.

    • [x] [prism] Next Step - Java Validation Tests - Unfilter tests from Gradle file.
      • beam/runners/portability/java/build.gradle
      • Did so locally and filed broad strokes failures above.
    • [ ] [prism] Next Step - Python Validation Tests - Unfilter skipped portable tests (super class)

Non-Go Blockers

Notable issues found in trying to run the Non Go SDKs (Java, Python, or others). Tracked in #28187, and more granular issues should be referred to here.

  • [ ] #28187
  • [x] Go SDK Cross Language PreCommit Suite (#28545)
    • [ ] No Basic Xlang Tests filtered. (#28544)
  • [ ] Prism Java Validates Runner Suite
    • [x] #31793
    • [ ] Gradle Target exists
    • [ ] Github Action exists
    • [ ] No tests filtered.
  • [ ] Python Validates Runner Suite
    • [x] #31583
    • [x] #32846
    • [x] #32847
    • [ ] No tests filtered.
  • [ ] Properly respect and handle SDK & Runner Capabilities.
  • [ ] #32564

Other Beam Core

This is an incomplete list of Beam features that would be nice to have.

  • [ ] Multi-Chunk Iterable protocol (#27762)
  • [ ] State Backed Iterables (for GBKs & Side Inputs)
    • [ ] State based Continuation Tokens
  • [ ] OrderedListState
  • [ ] #32139

Persistence & Reliability Features

Prism currently stores everything in memory. This includes all element data, in progress bundle data, pipeline info, artifacts etc. This is fast, but not the best use of memory for using prism long term as a stand alone runner.

  • [ ] Per Pipeline data should be moved to a local file cache.
    • They aren’t stored in memory when not needed. Eg. Artifacts shouldn’t live in memory once necessary environments are spun up.
    • [ ] Garbage collect artifacts after pipeline termination.
    • [ ] Garbage collect older pipelines after some threshold.
    • [ ] Separate Prism management logs and pipeline logs, with rolling log files
      • https://github.com/natefinch/lumberjack is a dependency free impl for this
  • [ ] Pipeline Restarts
    • [ ] Optimized stages need to be stored, so no complex mapping needs to occur for any persisted state.
    • [ ] Per stage pending elements and state needs to be stored so bundles can be re-computed on restarts.
      • It should be possible for a pipeline to be aborted, and prism torn down, and for a pipeline to be restarted from where it left off, with new worker processes.
      • FrostDB is an embeddable-in-Go, write optimized, in-memory + persistence, columnar database that might be a good thing to look at to enable these features.
  • [ ] Bundles Retries
    • Prism currently doesn’t retry failed bundles. A bundle failure fails the pipeline.
    • Adding a sensible retries policy would improve bundle reliability.
    • Affects how elements are divided into bundles, and scheduled.
    • Eg. A failed bundle could be split into smaller and smaller bundles, until the failing elements are isolated. Such a strategy would also enable implementation of error tolerance policies for example.
  • [ ] Improve (static) Bundle Splitting
    • Prism currently schedules all available pending elements into a single bundle.
    • Instead it could use some heuristic to determine how to split pending elements into new bundles to improve worker level parallelism before Channel or Sub Element Splitting occurs.
  • [ ] #32538
  • [ ] #29669
  • [ ] Pipeline Update
  • Similar to Cancel + Drain in combination with Pipeline restarts. Allow a pipeline to be updated mid execution.
  • [ ] #32167
    • Distinguishes between pipeline workers to avoid needing a new port for each instance.
    • Need a single "multiplexer" layer to route between the handlers for given jobs and workers.
  • [ ] #30703
    • SDK side logs via the Beam logging service to the runner should be available via the API. This may just require an SDK side change, or at least be a toggle on the prism stand alone binary.
  • [ ] #31971

Performance features

These are non-user facing Beam features that Dataflow implements. In order for Prism to serve the purpose of validating pipeline locally before production runner execution, these are required, to reduce worker side execution differences.

  • [ ] Side Input + State Cache
  • [ ] Elements on ProcessBundleRequest
  • [ ] Elements on ProcessBundleResponse
  • [ ] Autosharded keys
  • [x] #31628
  • [ ] Eagerly Bucket elements by Key + Window for GBKs
  • [ ] Intern user-keys, tags strings, byte arrays to reduce memory bloat during stateful pipeline execution. Go is garbage collected, not magic.

Stand Alone UI Based Features

These are features that are best tied to the ability to understand a job in the UI.

  • [ ] Data Sampling + plumbing to UI
  • [ ] Worker Status support + plumbing to UI
  • [ ] Runner side PubSub Transform (due to being a Beam built in)
  • [ ] Display of Optimized stages in UI
  • [ ] Display of Graph structure in UI
    • [ ] Interactivity with same.
  • [ ] Display of Job Logs in UI
    • ...and storage thereof in local cache.

Other features

The following are known issues/desires without a specific categorization at present.

  • [ ] Prism Per Job Configurability
    • Being able to toggle or set specific configurations using PipelineOptions or similar.
    • AKA the described Variants approach.
  • [ ] Additional runner side execution metrics
    • [ ] Count Splits per Transform
    • [ ] Count executed "Bundles" Executed.
    • [ ] "stage" execution time
    • [ ] Num elements per stage
      • [ ] Num Keys per stage (if stateful)
    • [ ] Histograms or Timeseries of the above?
  • [ ] Public Documentation of Prism
    • [x] #32562
    • [ ] Add prism to the Runner Capability Matrix
    • [x] #32601
    • [ ] Add Prism to Quickstarts and Examples
    • [ ] Blog Post about Prism

Completed Work

This section should be structured similarly to the Beam Compatibility Matrix for ease of transition to populating it there.

  • Environment Execution
    • [x] LOOPBACK/External
    • [x] Docker
  • Basics
    • [x] DoFns
    • [x] GBKS
    • [x] Windows
    • [x] Side Inputs
      • [x] Map
      • [x] Iterator
      • [x] #31339
  • Scaling
    • [x] Splittable DoFn support
    • [x] ProcessContinuation support
  • Performance
    • [x] Fusion

lostluck avatar Dec 07 '23 00:12 lostluck