> That query doesnt do what I want, because (shame on me) I have > multiple > docs with //row elements.
And do those row elements have RXAUI child elements that would match the value? --> YES > But just for testing I ran it and it performs about the same as the > cts:search case. > (4.5 sec) so it seems to be using indexes in that case. > Even that seems to be too slow for me where the result set is 8 > records > of about 100 bytes each. Does the query I gave you return 8 results or does it return a lot more because of your spurious row entries? ----> Just a few more, the other spurious records only account for a few > My real question here is one I'm trying to discover. And one that I > think many people are asking. > Can I get MarkLogic to perform like an RDBMS in the (hopefully rare) > cases where the data really is like RDB data ? You can definitely get Mark Logic to return 8 records in a fraction of a second, and faster still if the 8 records have their expanded tree cache records in memory. The fact that you're not seeing that makes me think something else is going on. ----> Something is fishy then, because even using indexes its taking > 4 seconds to return 8 rows. I'm trying the RecordLoader but I dont think its going to be practical. 5 hours later and its only 10% done. I dont think I am willing to wait the next 50 hours. Whereas loading the original 1G xml file with fragmentation set loaded in about 30 minutes. Looking at "top" on the server it seems there is a LOT more overhead involved mainly disk access, for creating separate documents then separate fragments. Its at pretty much 100% wait time but I'm only getting about 10 TPS from record loader. Loading 3 mil record will take way too long that way. ---------------------------------------- David A. Lee Senior Principal Software Engineer Epocrates, Inc. [email protected] 812-482-5224 _______________________________________________ General mailing list [email protected] http://xqzone.com/mailman/listinfo/general
