Skip to content

Ingestion is not ignoring unknown feature in streaming source #99

Description

@pradithya

Expected Behavior

Ingestion should ignore feature ID in FeatureRow that was not specified in Import spec

Current Behavior

Ingestion tried to ingest the unknown feature and throw following exception:

"transform":"Convert feature types","message":"Unknown feature myentity.none.unknown_feature, spec was not initialized","stackTrace":"java.lang.IllegalArgumentException: Unknown feature myentity.none.unknown_feature, spec was not initialized\n\tat com.google.common.base.Preconditions.checkArgument(Preconditions.java:141)\n\tat feast.ingestion.model.Specs.getFeatureSpec(Specs.java:148)\n\tat feast.ingestion.transform.fn.ConvertTypesDoFn.processElementImpl(ConvertTypesDoFn.java:44)\n\tat feast.ingestion.transform.fn.BaseFeatureDoFn.baseProcessElement(BaseFeatureDoFn.java:41)\n\tat feast.ingestion.transform.fn.ConvertTypesDoFn$DoFnInvoker.invokeProcessElement(Unknown Source)\n\tat org.apache.beam.runners.core.SimpleDoFnRunner.invokeProcessElement(SimpleDoFnRunner.java:275)\n\tat org.apache.beam.runners.core.SimpleDoFnRunner.processElement(SimpleDoFnRunner.java:240)\n\tat org.apache.beam.runners.dataflow.worker.SimpleParDoFn.processElement(SimpleParDoFn.java:325)\n\tat org.apache.beam.runners.dataflow.worker.util.common.worker.ParDoOperation.process(ParDoOperation.java:44)\n\tat org.apache.beam.runners.dataflow.worker.util.common.worker.OutputReceiver.process(OutputReceiver.java:49)\n\tat org.apache.beam.runners.dataflow.worker.SimpleParDoFn$1.output(SimpleParDoFn.java:272)\n\tat org.apache.beam.runners.core.SimpleDoFnRunner.outputWindowedValue(SimpleDoFnRunner.java:309)\n\tat org.apache.beam.runners.core.SimpleDoFnRunner.access$700(SimpleDoFnRunner.java:77)\n\tat org.apache.beam.runners.core.SimpleDoFnRunner$DoFnProcessContext.output(SimpleDoFnRunner.java:621)\n\tat org.apache.beam.sdk.transforms.DoFnOutputReceivers$WindowedContextOutputReceiver.output(DoFnOutputReceivers.java:71)\n\tat

Steps to reproduce

Run ingestion with streaming source (PubSub / Kafka) and publish FeatureRow with unknown feature inside the stream.

Metadata

Metadata

Assignees

Labels

No labels
No labels

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions