Details
-
Task
-
Status: Closed (View Workflow)
-
P2
-
Resolution: Done
-
None
-
-
Folijet Sprint 159
-
5
-
Folijet
-
Orchid (R1 2023)
Description
Currently, when a user cancels job DI continues to process chunks related to this job.
In RawMarcChunksKafkaHandler and StoredRecordChunksKafkaHandler get jobExecutionId from Kafka headers, check JobExecution status, if canceled or deleted (empty result) - skip processing of the chunk (DI_ERROR messages NOT needed) without deserializing (unpacking) the chunks.
Check impact on job progress tracking - add same logic in DataImportKafkaHandler (DI_COMPLETE and DI_ERROR messages). Check behavior of FlowControl, create separate task to resolve any issues.
This method will increase the performance of the DI and reduce the load on the CPU.
TestRail: Results
Attachments
Issue Links
- blocks
-
MODSOURMAN-934 Release v3.6.0 (R1 Orchid)
-
- Closed
-
- defines
-
UXPROD-3725 NFR: Data Import Technical, NFR, & Misc work (Orchid R1 2023)
-
- Closed
-
- is cloned by
-
MODSOURMAN-982 SPIKE: Investigate why chunks continue processing after DI job is completed
-
- Open
-
- relates to
-
MODSOURMAN-933 Prevent jobs from hanging when cancelled job is deleted before BG processing is finished
-
- Closed
-
-
MODDATAIMP-428 Cancelled data import jobs sometimes continue on to completion
-
- Closed
-
-
MODSOURMAN-940 DI Jobs stuck when selecting a wrong profile
-
- Closed
-
-
MODSOURMAN-956 Stop processing the job with incorrect profile
-
- Open
-
-
MODSOURMAN-970 Jobs run immediately after canceled job take excess time
-
- Open
-