abhishekagarwal87 commented on issue #13182: URL: https://github.com/apache/druid/issues/13182#issuecomment-1272263726
> Yes while looking at the issue and changes above, I had observed that the time buckets generation for time-granularity queries could be improved. One crude way of stopping time-grain generation going out of hand is to limit the amount of grains allowed to be generated by the query. A better but effort requiring way could be to derive the time-grains from data as we read it, instead of generating the time buckets first and then passing data through it. A similar problem happens on the ingestion side as well, when someone accidentally selects the wrong column for timestamp and kaboom :) I think your crude solution makes sense. The limit could be high enough that we know something is wrong with the query if a query hits that limit. And the limit is low enough that it stops an OOM. like 100,000 seems like a good number. what do you think? -- 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: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
