Hi team, I saw this weird behavior in my local datahub. Not sure if anyone encountered this behavior before or could share some idea. 🙇 After starting my local datahub with

./docker/dev-without-neo4j.sh

, I do a first ingestion using

datahub ingest -c ~/Desktop/recipe1.yml

(it ingests from json file with MCEs for Dataset), and then do the second ingestion using

datahub ingest -c ~/Desktop/recipe2.yml

(it ingests from json file with MCEs for Dashboard), The

getSearchResultsForMultiple

gql query with query variable “_types_” set to the first entity type (in this case, Dataset) return expected results. However, it returns no result when “_types_” is set to the second entity type (in this case, Dashboard). Supposedly, there should be some entities returned as Dashboard entities are ingested in the second run. I checked the ES, the entities exist in the

dashboardindex_v2

index. If I do the reverse, ingest Dashboards first then Datasets, I am not able to get Datasets from

getSearchResultsForMultiple

. Does this have anything to do with Elasticsearch cache?

Digesting 🙂

I don’t think it’s related to caching. Mainly because you used multiple filters to search. Do you still see this issue? It may take some time after the ingest command finishes for the search index to get updated

yup i think it just took some time for search to function probably! However, it’s flaky - occasionally it works immediately after ingest from CLI. Thanks John & Dexter, let me monitor for some while 🙇