September 1, 2016 at 10:36 pm
My company has an ecommerce application that can sometimes exceed 60,000 trans/sec. Things are being held together well enough with duct tape and baling wire but future success requires that increasing traffic. I am looking for some detailed case studies for similar applications with very high levels of transactional processing - eg 100,000+ trans/sec.
I have found some references in SQLCAT presentations about an unnamed securities trading organization that achieved 500,000 trans/sec peak with 350,000 trans/sec sustained.
I know the solution will involve some kind of scaling out but want to know how they did it. I can design a sharded system easily enough (relatively speaking) but what about how to scale easily from the initial setup. For example, how can more servers be added to accommodate additional load? How is the schema on the various shards updated as needed? How does a DBA manage and monitor such a creature?
I can probably fumble my way through the questions eventually with a lot a pain and false trails followed but would not object to a bit of hand holding through the scariest parts. In fact, it would be nice to know exactly which are the scary parts before I find myself in the middle of them.
If anyone can put me in contact with the DBA that are managing such systems or can provide links detailing the architecture, it would be much appreciated.
September 1, 2016 at 10:43 pm
You should contact MS directly. I'm sure they'd be glad to provide some information on this.
--Jeff Moden
Change is inevitable... Change for the better is not.
Viewing 2 posts - 1 through 1 (of 1 total)
You must be logged in to reply to this topic. Login to reply