History Repeats: MySQL, MongoDB, Percona, and Open Source

History is repeating again. MongoDB is breaking out of the niche into the mainstream, performance and instrumentation are terrible in specific cases, MongoDB isn’t able to fix all the problems alone, and an ecosystem is growing.


This should really be a series of blog posts, because there’s a book’s worth of things happening, but I’ll summarize instead. Randomly ordered:

  • MongoDB is in many respects closely following MySQL’s development, 10 years offset. Single index per query, MyISAM-like storage engine, etc. Background.
  • Tokutek built an excellent transactional storage engine and replaced MongoDB’s, calling it TokuMX. Results were dramatically better performance (plus ACID). MongoDB’s response was to buy WiredTiger and make it the default storage engine in MongoDB 3.0.
  • Percona acquired Tokutek. A book should be written about this someday. The impact to both the MySQL and MongoDB communities cannot be overstated. This changes everything. It also changes everything for Percona, which now has a truly differentiated product for both database offerings. This moves them solidly into being a product company, not just support/services/consulting; it is a good answer to the quandary of trying to keep up with the InnoDB engineers.
  • Facebook acquired Parse, which is probably one of the larger MongoDB installations.
  • Facebook’s Mark Callaghan, among others, stopped spending all his time on InnoDB mutexes and so forth. For the last year or so he’s been extremely active in the MongoDB community. The MongoDB community is lucky to have a genius of Mark’s caliber finding and solving problems. There are others, but if Mark Callaghan is working on your open source product in earnest, you’ve arrived.
  • Just as in MySQL, but even earlier, there are lots of -As-A-Service providers for MongoDB, and it’s likely a significant portion of future growth happens here.
  • MongoDB’s conference is jaw-droppingly expensive for a vendor, to the point of being exclusive. At the same time, MongoDB hasn’t quite recognized and embraced some of the things going on outside their walls. If you remember the events of 2009 in the MySQL community, Percona’s announcement of an alternative MongoDB conference might feel a little like deja vu. I’m not sure of the backstory behind this, though.

At the same time that history is repeating in the MongoDB world, a tremendous amount of stuff is happening quietly in other major communities too. Especially MySQL, but also in PostgreSQL, ElasticSearch, Cassandra and other opensource databases. I’m probably only qualified to write about the MySQL side of things; I’m pretty sure most people don’t know a lot of the interesting things that are going on behind the scenes that will have long-lasting effects. Maybe I’ll write about that someday.

In the meanwhile, I think we’re all in for an exciting ride as MongoDB proves me right.

PS: VividCortex is building a MongoDB monitoring solution that will address many of the shortcomings of existing ones. (We have been a bit quiet about it, just out of busyness rather than a desire for secrecy, but now you know.) It’s in beta now.

Cropped image by 96dpi

See Also

I'm Baron Schwartz, the founder and CEO of VividCortex. I am the author of High Performance MySQL and lots of open-source software for performance analysis, monitoring, and system administration. I contribute to various database communities such as Oracle, PostgreSQL, Redis and MongoDB. More about me.