So we are having an issue with `Explore from here`...
# gooddata-cloud
f
So we are having an issue with
Explore from here
. For users it just opens a blank white frame while loading. It then hangs for more than 2 minutes. After that it loads like nothing happened. The console log however returns this:
{status: 400, detail: A result cache error has occurred during the calculation of the result', resultId: 'a2b3ab1736e4df31873267202601067895a68b9f', reason: 'Query timeout occurred', traceId: 'c6bec71c898bb582e99144fa7a78fd0d'}detail: "A result cache error has occurred during the calculation of the result"reason: "Query timeout occurred"resultId: "a2b3ab1736e4df31873267202601067895a68b9f"status: 400traceId: "c6bec71c898bb582e99144fa7a78fd0d"
So what I think its trying to do is to load the chart using cache/preload but fails so it's recalculating? We need help on this - waiting for 2+ minutes for explore from here is seriously frustrating our customers.
j
Hi Filip, I can see the following error in our logs:
Execution of SQL statement was cancelled due to timeout
GD Cloud has a hard coded limit for SQL query execution of 30s. If the data values are not retrieved within the specified limit, the query will timeout. Usually, the root cause can vary, but in general, it could be slow datasource, a lot of data, complicated queries, stored data not optimised for performance, network latency, high load on the datasource, etc. If the amount of the data is not the cause or it is not possible to add a filter to the visualization, another possible cause could be the DB performance - e.g. high load (too many SQL queries in the queue) or the fact that the DB might need more resources or to be optimised for better performance. Just to be sure, could you please confirm, review and share the following: • The query execution times (from the database) when the filter loads were triggered. • In addition to the
Start time
,
Duration
and
SQL
, would it be possible to include and/or verify the query queue time as well as the queue priority? The idea here is to verify that the query queue time plus the execution/duration time are not longer than 30s on the database side and that the queries have the sufficient priority. • Are there any workload management rules or specific setup on the database side that might impact the query priority or the time it spends in the queue?