25

BBHoss # are mapreduce queries intended to be live or batch? When I get upwards of 10,000 things in a bucket, it takes too long to run the query, like 10 seconds

drev1 # BBHoss: bucket based queries are not recommended in production. the queries work best against a known list of bucket/key pairs

BBHoss # drev1: is that a new feature?

drev1 # which?

BBHoss # querying over bucket-key pairs

drev1 # no, the system was designed to accept {Bucket, Key} inputs. I believe the bucket based queries was added on top

BBHoss # drev1: so if i have a database full of items, how would i query them for an autocomplete-like dropdown box?

drev1 # I've been using ElasticSearch to search my documents with Riak providing the persistent storage

benblack # BBHoss: i'm not expert on the subject, but generally you would probably want to have a document that acted as an index so you could just retrieve it in a single call rather than iterating over your entire bucket every time

BBHoss # benblack: so a single objects with 10s of thousands of hashes?

benblack # quite possibly, yes

rtilder joined #riak

scootsm # BBHoss: If you look at the pbclient I'm working on, you can list_buckets / list_keys.

BBHoss: Debating how to use it, while being mindful of performance. Does listing the keys inside a bucket cause a performance hit? well, obviously some Err, that wasn't a question to you, BBHoss. Just a general one

BBHoss # scootsm: yeah i think thats really expensive

scootsm # listing the keys would allow you to build a cache and query that way.

BBHoss # i think with riak, you are better off maintaining your own "caches" with links or keys of documents

scootsm # yeah, I think so. Although, listing keys (a lot of em) returns seemingly faster than listing buckets