https://datahubproject.io logo
Join Slack
Powered by
Questioning my sanity here, but hopefully someone ...
# troubleshoot
h
Questioning my sanity here, but hopefully someone can help clear things up. I have a new datahub deployment (version 0.8.11 ish), and did a test ingesting a dummy user. It seems to work, as I see 
c.l.m.r.entity.EntityResource - INGEST urn urn:li:corpuser:test-user with system metadata {lastObserved=1635959902784}
in the GMS log, and it shows up in search and in the UI, but I don't see it in the GMS database (
select * from metadata_aspect_v2
). The DB only contains the default 
urn:li:corpuser:datahub
. Am I missing something? 😅
g
Hey there @high-hospital-85984 - if the entity is coming up in the UI, that means there has to be a record in your db. I would double check: a) are you querying the right mysql instance? b) is it possibly a pagination issue? What if you issue a select star 
where urn=<your user urn>
oh, just to clarify- when the record appears in the UI, does it have its fields like name & description filled in?
or is it a skeleton entity?
If it just exists in elastic but not mysql, it will look like a skeleton entity showing just the information contained in the Urn
h
I think it's just a skeleton. No email showing, for example.
g
Mhm, interesting
can you try again?
ingesting a new user?
and see if that works?
h
I'm suspecting a connection problem to the database. We've been struggling with them a bit.
sure, one sec
Okey, tried a dataset this time. From the GMS logs:
Copy code
18:51:27.003 [qtp544724190-154] INFO  c.l.m.r.entity.EntityResource - INGEST urn urn:li:dataset:(urn:li:dataPlatform:temporary,test-dataset,PROD) with system metadata {lastObserved=1635965487003}
18:51:27.114 [pool-9-thread-1] INFO  c.l.metadata.filter.LoggingFilter - POST /entities?action=ingest - ingest - 200 - 111ms
Still nothing in the table. It shows up in the UI, but the description is missing.
g
do you have any logs from your database?
I've never seen this before 🤔
h
Sorry no, not with my rights, at this moment. I'll continue investigating tomorrow, when I can get some better access to the DB
g
alright, sounds good
let me know how it goes
👍 1
h
Okey, it seems to have solved itself with a restart of the GMS pod. As I mentioned before, we were going back and forth with the permission on the user, and even though we got rid of the errors in the logs, the service might have gotten stuck in some strange state. Will investigate more tomorrow
yup, all looking good now! Sorry for the confusion 😅
2 Views
Open in Slack
PreviousNext
Powered by