[ https://issues.apache.org/jira/browse/CASSANDRA-12915?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15742418#comment-15742418 ]
Corentin Chary commented on CASSANDRA-12915: -------------------------------------------- To give you some perspective, in my specific use case the queries are user generated and our only safeguard is that we put a limit to the number of generate results (5000). We already process user-provided globs into CQL queries trying to choose the best indexes. But really it would be better if the query optimizer was in the database rather than in our code. I agree that my way of doing things tries to bypass a few things, but it also doesn't seem to create regressions and provides some benefits. I propose that we split the discussion in three parts: 1# Do you see a quick way of fixing the 'empty iterator' issue ? 2# Would you agree to chance the format to include proper cardinality estimation to the index (so we can build upon later) 3# What would be the ETA for a proper query planner ? If O(months), would it make sense to merge something like what I did to bring some performance improvement in the meantime ? Thanks ! > SASI: Index intersection can be very inefficient > ------------------------------------------------ > > Key: CASSANDRA-12915 > URL: https://issues.apache.org/jira/browse/CASSANDRA-12915 > Project: Cassandra > Issue Type: Improvement > Components: sasi > Reporter: Corentin Chary > Fix For: 3.x > > > It looks like RangeIntersectionIterator.java and be pretty inefficient in > some cases. Let's take the following query: > SELECT data FROM table WHERE index1 = 'foo' AND index2 = 'bar'; > In this case: > * index1 = 'foo' will match 2 items > * index2 = 'bar' will match ~300k items > On my setup, the query will take ~1 sec, most of the time being spent in > disk.TokenTree.getTokenAt(). > if I patch RangeIntersectionIterator so that it doesn't try to do the > intersection (and effectively only use 'index1') the query will run in a few > tenth of milliseconds. > I see multiple solutions for that: > * Add a static thresold to avoid the use of the index for the intersection > when we know it will be slow. Probably when the range size factor is very > small and the range size is big. > * CASSANDRA-10765 -- This message was sent by Atlassian JIRA (v6.3.4#6332)