Poor performance with B-Trees

[Joan Miller]

What persistence model has mongodb? If it's B-trees:

I'm supposed that you know:

"B-Trees have been used in databases since practically the beginning
of time. They provide robust indexing support, but performance is
poor on rotational disks (which are still by far the most cost-
effective) because of the multiple seeks involved in reading or
writing anything."

And one better alternative would be memtables and SSTables

http://www.rackspacecloud.com/blog/2009/11/09/nosql-ecosystem/

[Dwight Merriman]

Actually b-trees are great.  B-Trees were original designed for rotational disks.

One reason BigTable uses an alternate mechanism: it runs atop a grid file system (GFS).  That system allows only writes of large blocks at a time.  Thus, they needed a design which never does granular writes.  So my guess is the sstable approach was more from desire for running on GFS than to speed inserts per se (although there may be benefits there).

So in general, we are very happy with mongodb btree performance.  If you want to make a specific benchmark vs. something that uses something else we would be happy to look at it afterward and dive deeper.

Also interest that would is moving more towards solid state drives too, adding even more nuance to database performance!

--

You received this message because you are subscribed to the Google Groups "mongodb-user" group.
To post to this group, send email to mongod...@googlegroups.com.
To unsubscribe from this group, send email to mongodb-user...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/mongodb-user?hl=en.