gaborkaszab commented on PR #16574: URL: https://github.com/apache/iceberg/pull/16574#issuecomment-5088783749
Hi @moomindani , Would be nice to hear the opinion of experts on this, but here is why I'm somewhat hesitant on the approach: 1. The PR defines table name regexes that we include / exclude from metrics reporting. This assumes that we can decide whether or not to report metrics for a table by applying some predicate on its name. Are we sure that name is enough to judge this? What if we introduce a regex now, but then it excludes some future tables that we don't want to exclude? I don't see how exactly this functionality would be used in prod. Maybe having a single list of table names that we want to exclude and expect exact match instead of regex? 2. These regexes are configured via catalog properties. When we want to change them (for instance we have a new table that matches some of the regexes we don't want) then this requires restarting the client-side catalog to pick up new values, right? There might be engines that start a new client-side catalog connector for each query, for them this isn't an issue. I know of engines that launch the client-side catalog connector when the engine launches and keeps this single instance open. For them this design might be rigid and inconvenient. Before moving forward I'd suggest having a wider community opinion on this. -- 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]
