Hello, team. After some ingestion/deleting procedu...
# troubleshoot
c
Hello, team. After some ingestion/deleting procedures i can't find dbt datasets via search and browse paths. Only by direct url with urn. For other platforms all works fine. Elastic re-index dont help. What should i can try to fix it?
For re-index using this
Copy code
kubectl create job --from=cronjob/datahub-datahub-restore-indices-job-template datahub-restore-indices-adhoc
b
most probably the deleted entities have the removed:true Status aspect. (Not sure which version of datahub you have, the latest version have a banner on the profile to indicate that the dataset is soft deleted)
does reingesting help to reactivate them?
c
Copy code
does reingesting help to reactivate them?
no, it doesn't
b
is there alot of datasets affected? wondering if you should try the rollback or ingest a status aspect. not sure why only dbt is affected
c
Yes, a lot. I did rollback and now i can see some datasets (not all). And in list-runs there are some runs that i cant delete
No entities found. Payload used: {blabla}
Hard delete by urn and reingest fix my problem )
Copy code
datahub delete --urn "blabla" --hard
Is there any way to hard delete all platform entities? This don't work, it delete only "visible" datasets
Copy code
datahub delete --platform dbt --hard
b
Doesn't seem possible. I would either ingest status aspect to make them not soft deleted, or go into MySQL to delete the rows and reindex ES
thank you 1
c
go into MySQL to delete the rows and reindex ES
i did it and it didn't help
ingest status aspect
Do you have any example?
b
https://datahubproject.io/docs/metadata-service/ under section soft deleting an entity
I'm surprised deleting in mysql didn't work though
Are you deleting them so that they can be subsequently re ingested?
c
I think reindex don't work. Because now i have some datasets visible in UI which don't have any info in mysql. After reindex same result.
b
could you share more on where those "doesnt exist in mysql" datasets are displayed? in the Datasets view page? inside a dataset lineage tab?
c
some screenshots
b
oooh, interesting. No idea how this could have happened. ^ @big-carpet-38439
c
how i do reindex:
Copy code
C:\Users\ntsyplakov>kubectl --namespace datahub create job --from=cronjob/datahub-datahub-restore-indices-job-template datahub-restore-indices-adhoc
job.batch/datahub-restore-indices-adhoc created
C:\Users\ntsyplakov>kubectl --namespace datahub get pod
NAME                                               READY   STATUS      RESTARTS   AGE
...
datahub-restore-indices-adhoc-jd4wr                0/1     Completed   0          28m
logs
Copy code
Starting upgrade with id RestoreIndices...
Cleanup has not been requested.
Skipping Step 1/3: ClearSearchServiceStep...
Cleanup has not been requested.
Skipping Step 2/3: ClearGraphServiceStep...
Executing Step 3/3: SendMAEStep...
...
Successfully sent MAEs for 16797 rows
Completed Step 3/3: SendMAEStep successfully.
Success! Completed upgrade with id RestoreIndices successfully.
Upgrade RestoreIndices completed with result SUCCEEDED. Exiting...
b
that should have been the correct way to reindex
thank you 1