You can always tell when someone’s showing off their cool, cutting edge comms technology. They tend to raise their voice a lot.
Back in the day they’d announce their gadget leadership to the rest of the herd by shouting into their cellphone. Usually the message was no more urgent than “Hi, I’m on my cellphone!” Now the same types will loudly name-drop a different technology to the rest of the airport lounge. “I’m leveraging the wifi,” a fellow passenger bellowed, the other day, as we filtered through the departure gate. Nobody needed to know that, but the subtext was “look at me everybody”.
You can tell the really advanced mobile user – they tend to whisper. Their handset has a microphone (how cool is that!) and they know how to use it.
Sometimes these shouty public broadcasters aren’t even connected anyway because the database for their Voice over IP (VoIP) platform can’t cope. This will happen if they are using a traditional SQL model to try and cope with a phone network which has far flung offices and hundreds of mobile employees. That, like shouting into your phone, is just wrong on so many levels.
What VoIP needs now is a single, logical database across multiple servers in different geographies. It needs to be updated in real-time and automatically scaled out during times of peak demand. A VoIP system should scale up to handle increased traffic, but just as importantly is must then go back down in the off peak hours.
Try this with a MySQL database. It can’t scale easily enough, so it will keep your developers busy. They’ll have spent many hours trying to knit the different databases together.
Traditional relational databases can possibly achieve this, at a price. Mind you, you could extend baked bean cans and string to every point on the network and that would be no less elegant. That’s not really following engineering principles though is it?
Having said that, most telcos and VoIP systems use a separate, independent solution for each office location, which they link together – loosely. The more office locations, the more complex and expensive the solution becomes and so the more you spend on maintenance. Ideally, you’d have a fluid system that can automatically shift its shape as the need arises. That’s the point of software isn’t it – it adapts. Otherwise, we might as well return to the old days. A MySQL system isn’t exactly baked bean cans attached by string, but it’s closer in spirit to the old many teethed mechanical beast that was employed in the first type of automated switchboard.
NuoBD’s NewSQL is designed to be a single database that works across multiple servers, which can scale easily, and scale on demand. That’s one system that gives high connectivity but no latency, complexity or maintenance issues.
MySQL works in some circumstances, but a period of growth isn’t one of them. So as a company moves forward, the MySQL database can’t keep pace. Data storage and data replication errors creep in. Soon the diaspora of offices becomes a problem. Your telephone system isn’t just distributed, it is literally all over the place.
Though voice calls are often a software function, some of the old habits of telephony remain. When you call an engineer out, some of them will listen to what you’re asking for and announce that it cannot be done. This is what happens if you ask, say, database engineers familiar with Oracle or Microsoft to fulfill your wish for a low maintenance system built on a single, fluid, scalable database. No can do, they’d say. In fact, I heard one shouting something similar into his VoIP handset at the airport. “I can’t get on the network, Mac. I’m on MySQL.”
You can download NuoDB from here.
“NuoDB provides the ability to replicate data globally in real-time, which is not available with any other product offering,” states Weeks. “That alone is remarkable and it works. I’ve seen it. I’ve used it. I’ve tested it. The ability to deploy NuoDB removes a tremendous burden from our support and engineering teams.”
Filed under: PostADay, SQL, SQL Authority, SQL Query, SQL Server, SQL Tips and Tricks, T SQL Tagged: NuoDB
