christian-schlichtherle opened a new issue, #15224: URL: https://github.com/apache/druid/issues/15224
I’m having trouble using window functions in Druid v27.0.0. I have added `"windowsAreForClosers": true` to the query context. For demonstration purposes, I am using the data source described in this blog posting: https://blog.hellmar-becker.de/2023/03/26/druid-26-sneak-peek-window-functions/ Here’s my query: ```sql with t1 as ( select floor(__time to day) as "date", array_agg(distinct user_id) as user_ids from events group by 1 ) select "date", array_concat_agg(distinct user_ids, 65536) over (order by "date" rows unbounded preceding) as cumulative_user_ids from t1 order by 1 ``` This is the response from the web console: ``` Error: UNCATEGORIZED (ADMIN) Error while applying rule DruidOuterQueryRel(WINDOW), args [rel#61203:LogicalWindow.NONE.[](input=RelSubset#61183,window#0=window(partition {} order by [0] rows between UNBOUNDED PRECEDING and CURRENT ROW aggs [ARRAY_CONCAT_AGG($1, $2)])), rel#61214:DruidQueryRel.NONE.[0](query={"queryType":"timeseries","dataSource":{"type":"table","name":"events"},"intervals":{"type":"intervals","intervals":["-146136543-09-08T08:23:32.096Z/146140482-04-24T15:36:27.903Z"]},"granularity":"DAY","aggregations":[{"type":"expression","name":"a0","fields":["user_id"],"accumulatorIdentifier":"__acc","initialValue":"ARRAY<LONG>[]","initialCombineValue":"ARRAY<LONG>[]","isNullUnlessAggregated":true,"shouldAggregateNullInputs":true,"shouldCombineAggregateNullInputs":false,"fold":"array_set_add(\"__acc\", \"user_id\")","combine":"array_set_add_all(\"__acc\", \"a0\")","maxSizeBytes":1024}],"context":{"queryId":"0caf2e20-cfc1-44d7-9903-4125bc08ac6e","skipEmptyBuckets":true,"sqlOuterLimit":1001,"sqlQueryId":"0c af2e20-cfc1-44d7-9903-4125bc08ac6e","timestampResultField":"d0","windowsAreForClosers":true}},signature={d0:LONG, a0:ARRAY<LONG>})] ``` For comparison, this works (although not very useful): ```sql with t1 as ( select floor(__time to day) as "date", count(distinct user_id) as user_ids from events group by 1 ) select "date", sum(user_ids) over (order by "date" rows unbounded preceding) as cumulative_user_ids from t1 order by 1 ``` Please include as much detailed information about the problem as possible. - Cluster size: One node for development - Configurations in use: `druid.generic.useDefaultValueForNull=false` -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: commits-unsubscr...@druid.apache.org.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: commits-unsubscr...@druid.apache.org For additional commands, e-mail: commits-h...@druid.apache.org
