Hi there, I'm facing an error with the user roles....
# troubleshoot
s
Hi there, I'm facing an error with the user roles. I'm in version v.0.9.3 and we have SSO authentication. So, the problem is that if I change the role to a new user (by default i got "No Role") to "Editor" or "Reader" and then I refresh the page, the new role is gone. Again the user will appear with "No Role". Any suggestion?? Thanks
đź‘€ 1
a
Hi Adriana, could you post some screenshots of the behavior? What part of the UI are you changing them in?
b
Is this happening every time, or sporadic?
I do recall us submitting a fix for a similar behavior that was non-deterministic, it wasn't always happening
s
The behaviour is happening every time, it's not sporadic
what I do is to go to the UI "Settings">"User&Groups" and then on the right part of the name of the user there is a list box where I select the role.
for ex if in this case I select "Editor" and then I refresh the page I got again "No Role"
a
Hi @steep-midnight-37232, this should be fixed 0.9.5, is this issue still affecting you?
s
Ah, ok. let me update the version then and check. Thanks. I'll let you know
Hi Paul, I have tried to upgrade but got this error "Field kafkaHealthChecker in com.linkedin.gms.factory.kafka.DataHubKafkaEventProducerFactory required a bean of type 'com.linkedin.metadata.dao.producer.KafkaHealthChecker' that could not be found." in the datahub-upgrade-job. Any suggestion?
thanks
Hey guys the upgrade succeeded with 0.9.5.1rc2 but the problem with the roles is still there.
a
Thanks for the update @steep-midnight-37232, we’ll try and get this figured out for you! CC: @big-carpet-38439
s
Hey any news? Still with the same problem. Thanks
b
Hi @steep-midnight-37232 this is definitely a bug that we fixed. Let's me find the exact PR and release where it is there
Technically it means this fix should have been in v0.9.3, unless this is a completely separate issue
cc @echoing-airport-49548
It would be useful to get on a call
Having a hard time reproducing on latest
Aha ! I managed to catch this happening one time
Okay let me look.
e
Hey @steep-midnight-37232 what are you using as your graph service? neo4j or elastic?
@steep-midnight-37232 can you also tell me what exception you’re seeing on the GMS pod when you try to assign the role?
s
Hey guys, thanks for your suggestions. We have both install neo4j and elastic as by default.
We also noticed that it appeared after looker ingestion ... not sure if it's connected to this
this is the log:
073039.380 [pool-11-thread-1] WARN org.elasticsearch.client.RestClient:65 - request [POST http://elasticsearch-master:9200/datahubpolicyindex_v2/_search?typed_keys=true&max_concurrent_shard_requests=5&ignore_unavailable=false&expand_wildcards=open&allow_no_indices=true&ignore_throttled=true&search_type=query_then_fetch&batched_reduce_size=512&ccs_minimize_roundtrips=true] returned 2 warnings: [299 Elasticsearch-7.16.2-2b937c44140b6559905130a8650c64dbd0879cfb "Elasticsearch built-in security features are not enabled. Without authentication, your cluster could be accessible to anyone. See https://www.elastic.co/guide/en/elasticsearch/reference/7.16/security-minimal-setup.html to enable security."],[299 Elasticsearch-7.16.2-2b937c44140b6559905130a8650c64dbd0879cfb "[ignore_throttled] parameter is deprecated because frozen indices have been deprecated. Consider cold or frozen tiers in place of frozen indices."] 073159.707 [ThreadPoolTaskExecutor-1] INFO c.l.m.k.t.DataHubUsageEventTransformer:74 - Invalid event type: SelectUserRoleEvent 073159.707 [ThreadPoolTaskExecutor-1] WARN c.l.m.k.DataHubUsageEventsProcessor:56 - Failed to apply usage events transform to record: {"type":"SelectUserRoleEvent","roleUrn":"urnlidataHubRole:Admin","userUrn":"urnlicorpuser:pau.fores","actorUrn":"urnlicorpuser:datahub","timestamp":1674804720255,"date":"Fri Jan 27 2023 083200 GMT+0100 (Central European Standard Time)","userAgent":"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/108.0.0.0 Safari/537.36","browserId":"03a07a18-29a3-4bf4-8c09-41c8b7bb89d3"} 073203.219 [ForkJoinPool.commonPool-worker-835] WARN org.elasticsearch.client.RestClient:65 - request [POST http://elasticsearch-master:9200/datahubroleindex_v2/_search?typed_keys=true&max_concurrent_shard_requests=5&ignore_unavailable=false&expand_wildcards=open&allow_no_indices=true&ignore_throttled=true&search_type=query_then_fetch&batched_reduce_size=512&ccs_minimize_roundtrips=true] returned 2 warnings: [299 Elasticsearch-7.16.2-2b937c44140b6559905130a8650c64dbd0879cfb "Elasticsearch built-in security features are not enabled. Without authentication, your cluster could be accessible to anyone. See https://www.elastic.co/guide/en/elasticsearch/reference/7.16/security-minimal-setup.html to enable security."],[299 Elasticsearch-7.16.2-2b937c44140b6559905130a8650c64dbd0879cfb "[ignore_throttled] parameter is deprecated because frozen indices have been deprecated. Consider cold or frozen tiers in place of frozen indices."] 073203.222 [ForkJoinPool.commonPool-worker-379] WARN org.elasticsearch.client.RestClient:65 - request [POST http://elasticsearch-master:9200/corpuserindex_v2/_search?typed_keys=true&max_concurrent_shard_requests=5&ignore_unavailable=false&expand_wildcards=open&allow_no_indices=true&ignore_throttled=true&search_type=query_then_fetch&batched_reduce_size=512&ccs_minimize_roundtrips=true] returned 2 warnings: [299 Elasticsearch-7.16.2-2b937c44140b6559905130a8650c64dbd0879cfb "Elasticsearch built-in security features are not enabled. Without authentication, your cluster could be accessible to anyone. See https://www.elastic.co/guide/en/elasticsearch/reference/7.16/security-minimal-setup.html to enable security."],[299 Elasticsearch-7.16.2-2b937c44140b6559905130a8650c64dbd0879cfb "[ignore_throttled] parameter is deprecated because frozen indices have been deprecated. Consider cold or frozen tiers in place of frozen indices."]
e
This is from GMS right?
you’re not seeing anything about how it failed to assign the role?
s
This is the error: Invalid event type: SelectUserRoleEvent WARN c.l.m.k.DataHubUsageEventsProcessor:56 - Failed to apply usage events transform to record
e
Hey sorry for not following up sooner, but there were no other errors?
s
nope, just this
e
Okay if you’re using Neo4j, there are known issues there, so I would expect that to be the issue, unfortunately
If it’s Elastic we may need to have a debugging session, if you’re open to it
s
Hey Aditya, I have just update to version 0.10 and the problem seems solved.
Thank you very much for the help