tonysun83 commented on issue #471: Use efficient set storage for field names URL: https://github.com/apache/couchdb/pull/471#issuecomment-292075085 @sagelywizard : yes, it does just push the failure over to Clouseau. We've already seen in production when users index everything, and the number of unique fields blows up the heap memory. In this particular case, there were so many fields that it never even made it to clouseau and got stuck in this process. We have an option for users to disable the indexing of everything in the document, but if they still decide index a heavily nested field with objects and arrays, then this issue could still occur. A while back, we discussed the possibility of adding another limit (# of levels deep we go in the object, or some total number of fields). Unfortunately, that would mean we'd also lose some ability for querying because those fields won't be indexed. That's for a bigger PR. This is more of a cosmetic change, so we don't necessarily have to merge it. ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected]
With regards, Apache Git Services
