errors in executor log - Can't save pipeline messages

[martina-parkanova]

We have a problem - but only on one of our environments - the same pipeline runs fine on one environment, but on another it finishes after "JSON to JSON-LD" component and even though it is marked green, it doesn't continue its execution and the next component is still with status "waiting to be executed". In the execution logs, there is a strange error which we don't understand, could you please tell us what could we do with it?

2024-05-17 15:44:30,924 [Files to RDF single graph] INFO c.l.e.e.p.v.s.DefaultProgressReport - Progress report ALL/1
2024-05-17 15:44:30,925 [Files to RDF single graph] INFO c.l.e.e.e.ExecutionObserver - onComponentUserCodeSuccessful : http://localhost:8080/resources/pipelines/1715850292163-00021715850292163-00021144-92ee
2024-05-17 15:44:30,925 [asynchExecutor-1] INFO c.l.e.e.e.ExecutionObserver - onExecuteComponentSuccessful : http://localhost:8080/resources/pipelines/1715850292163-00021715850292163-00021144-92ee
2024-05-17 15:44:30,930 [asynchExecutor-1] DEBUG c.l.e.d.c.r.RepositoryManager - Request to close repository group: 'single' used: '1'
2024-05-17 15:44:30,931 [asynchExecutor-1] INFO c.l.e.e.e.ExecutionObserver - onExecuteComponentInitializing : http://localhost:8080/resources/pipelines/1715850292163-00021715850292163-00021714045949842-00011714045949842-00011707992086526-00011707992086526-00011701862950169-00011701862950169-00011700478758489-00011700478758489-00011699965851915-00011699965851915-00011699883818893-00031699883818893-00037d97-acf9
2024-05-17 15:44:30,935 [asynchExecutor-1] ERROR c.l.e.e.e.ExecutionObserver - Can't save pipeline messages.
java.io.FileNotFoundException: /data/lp-etl/executor/data/1715960570672-20-6c8a992d-c54f-4df1-8b6f-e49b158c96c1/messages/1715850292163-00021715850292163-00021714045949842-00011714045949842-00011707992086526-00011707992086526-00011701862950169-00011701862950169-00011700478758489-00011700478758489-00011699965851915-00011699965851915-00011699883818893-00031699883818893-00037d97-acf9.trig.swp (File name too long)
at java.base/java.io.FileOutputStream.open0(Native Method)
at java.base/java.io.FileOutputStream.open(FileOutputStream.java:293)
at java.base/java.io.FileOutputStream.(FileOutputStream.java:235)
at java.base/java.io.FileOutputStream.(FileOutputStream.java:184)
at com.linkedpipes.etl.rdf.rdf4j.Rdf4jUtils.save(Rdf4jUtils.java:40)
at com.linkedpipes.etl.executor.execution.message.BaseMessageWriter.save(BaseMessageWriter.java:79)
at com.linkedpipes.etl.executor.execution.message.ComponentMessageWriter.save(ComponentMessageWriter.java:111)
at com.linkedpipes.etl.executor.execution.ExecutionObserver.writeComponentMessagesToDisk(ExecutionObserver.java:159)
at com.linkedpipes.etl.executor.execution.ExecutionObserver.onExecuteComponentInitializing(ExecutionObserver.java:201)
at com.linkedpipes.etl.executor.component.ExecuteComponent.execute(ExecuteComponent.java:71)
at com.linkedpipes.etl.executor.pipeline.PipelineExecutor.executeComponent(PipelineExecutor.java:346)
at com.linkedpipes.etl.executor.pipeline.PipelineExecutor.executeComponents(PipelineExecutor.java:319)
at com.linkedpipes.etl.executor.pipeline.PipelineExecutor.execute(PipelineExecutor.java:95)
at com.linkedpipes.etl.executor.web.servlet.ExecutionServlet.lambda$execute$0(ExecutionServlet.java:72)
at java.base/java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1136)
at java.base/java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:635)
at java.base/java.lang.Thread.run(Thread.java:833)

[martina-parkanova]

So I fixed that problem, but it was a strange fix. I just tried to delete the json to json-LD component and create it manually again (on this environment, the pipeline was imported).
So I guess there is some bug in export-import when the link to the next component is not persisted in some case... I don't have the data now, but I can save the corrupted pipeline next time when it happens if you want.

[skodapetr]

Dear Martina, sharing the corrupted pipeline would be definitely beneficial and hopefully would help us prevent such errors in the future.

[martina-parkanova]

Hello,
no problem - it is actually very easy to simulate this problem. You just need two or three components, I have these three, but I think the main problem is the JSON to JSON-LD component:

When you have such pipeline, you can simply copy the whole pipeline in the main pipeline list and the copied pipeline is corrupted - runs the JSON to JSON-LD component with success, but doesn't continue.
I attach my simplified pipeline having this problem.
corrupted-pipeline.zip

Btw - different problem, but I returned from the component "JSON-LD to RDF Titanium" back to "Files to RDF single graph" because the Titanium one seems like it doesn't support UTF-8 and I had problems with the output text format.

have a nice day,
Martina