Hi, do you know where I can find some basic tutor...
# getting-started
s
Hi, do you know where I can find some basic tutorial? As example, I successfully added metadata from my SQL Server source, but I don't know how to delete the test metadata ingested from the "quickstart"
e
Hi @stale-jewelry-2440. Deleting metadata is not yet supported. We plan to support it in the near future, as it is a very high priority item for us. In the meantime, can you try starting fresh by running
/docker/nuke.sh
and then reingesting the SQL source?
s
trying now...
e
If you are blocked by anything else, please let us know. For ingestion https://datahubproject.io/docs/metadata-ingestion is a great place to start!
docker/nuke stops all containers, so you will have to run
/docker/quickstart.sh
again before ingesting!
s
it seems some more step is needed: launching
./docker/ingestion/ingestion.sh
i get an error:
ERROR: for ingestionĀ Cannot start service ingestion: network xxxxx not found
e
Can you run
docker system prune
?
s
maybe I did something equivalent, by adding
--remove_orphans
into the
/docker/quickstart.sh
script ingestion gone fine
e
Yes. Seems like
nuke.sh
does not delete the old ingestion container and the network it was referring to. Sending out a PR for this now.
šŸ‘ 2
s
thnk you!
b
Thanks Dexter!
c
@stale-jewelry-2440 if you just don't want those datasets to show up in DataHub UI, you probably can set those test ones to 'removed' status by ingesting them again with extra status aspect. That works for me for Chart entity. I haven't tried it with datasets, but assuming that should work similarly.
e
Sorry about the late msg. Here is the PR https://github.com/linkedin/datahub/pull/2459
m
@stale-jewelry-2440 did those steps work to clear out the initial metadata?
s
Hi! I did not tried with the "remove" flag, just the "nuke" procedure and after a minor manual action, described above, it worked nicely. Thank you!
šŸŽ‰ 1