For some reason I can no longer read a specific directory in HDFS. None of these folders have formats applied to them even though they contain jsonl.gz files…
com.dremio.service.namespace.NamespaceNotFoundException: one or more elements on the path are not found in namespace: “HDFS”.proj.hld.raw.“year=2018”
Dremio says that for /proj/hld/raw/year=2018 I have one dataset, but I’m not sure why it is there. I can’t find an equivalent virtual data set for it.
Trying to access the monthly folders by adding them to the URL works except for month=02
com.dremio.service.namespace.NamespaceNotFoundException: one or more elements on the path are not found in namespace: “HDFS”.proj.hld.raw.“year=2018”.“month=02”
/proj/hld/raw/year=2018/month=01 is accessible
/proj/hld/raw/year=2018/month=02 is not accessible
/proj/hld/raw/year=2018/month=03 is accessible
/proj/hld/raw/year=2018/month=04 is accessible
Even if I move month=02 to /tmp, the parent folder year=2018 still returns the error above.
I even removed the entire year=2018 folder and manually created an empty year=2018 folder or year=2018/month=01 with some valid files in it and the entire 2018 folder still errors out.
I’ve tried running both:
ALTER PDS “HDFS HALFARM”.proj.mstar.hld.raw.“year=2018”.“month=02” REFRESH METADATA
ALTER PDS “HDFS HALFARM”.proj.mstar.hld.raw.“year=2018” REFRESH METADATA
Unable to find requested table “HDFS HALFARM”.proj.mstar.hld.raw.“year=2018”.“month=02”
Unable to find requested table “HDFS HALFARM”.proj.mstar.hld.raw.“year=2018”