Testing BlitzDB on Drizzle’s Build Farm
One of many important things that the Drizzle project takes seriously is for the project sourcecode to successfully build in all our target platforms AND pass tests in them. This is not really specific to Drizzle as most open source projects would have the same policy. For example we do the same thing in memcached thanks to Dustin Sailing’s buildbot kungfu.
Yesterday, Monty Taylor gave me access to Drizzle’s Build Farm Infrastructure so that I could test BlitzDB on various Linux distributions and FreeBSD. Unfortunately most build machines didn’t have Tokyo Cabinet installed so I could only test builds on Ubuntu and Debian. Fortunately the build went fine on those platforms though this was predictable since Ubuntu is my primary development platform. What was disturbing was getting test errors on my index test suite. I guess it’s time to put my thinking cap on and see what the problem is there.
This is a big leap towards getting BlitzDB in Drizzle’s trunk which I’m steadily working towards. I also want to benchmark BlitzDB at it’s current state with sysbench‘s OLTP tests. This is still low in my priority queue but hopefully I’ll do it in the next couple of months.
Toru,
Are you also testing BlitzDB with the random query generator? If not, I highly recommend it – it has proven to be an invaluable tool in testing MySQL
@patrick Hi Patrick. Thanks for the advise. Admittedly, I don’t know many tools for testing MySQL and Drizzle so I really appreciate advises in this area.
Am I correct to say that this is the tool that you recommend?
– https://launchpad.net/randgen
Currently I plan on testing BlitzDB with sysbench, tpce-mysql, drizzleslap (mysqlslap) and skyload which is a small non-blocking parallel benchmark program I wrote a while back.
Hi. Yes, that was exactly what I meant : )
I’m not 100% certain how well it will play with BlitzDB at the moment, but the randgen team is very interested in having lots of people use the tool and will be open to making things work well for everyone.
I mainly like the RQG as it can generate a lot of pretty mean queries fairly quickly. You can expect me to start writing more about using this tool in a few weeks, so we might be able to collaborate on this then. I’m interested in trying to beat up BlitzDB with it at any rate ; )
@patrick Sounds good! More tests the merrier.
I first thought RQG may not play well with Drizzle due to reduced number of types but I’ve talked to other Drizzle hackers and it seems we’ve ported RQG to work with Drizzle. This is exciting.
Would you be able to give me the URL to where you’re planning to write about this? Something I’m interested is writing an import feature to skyload that accepts an RQG output. I already wrote an import feature (though it’s still premature) so it shouldn’t be too tough to pull off.