Remember last year when the the Mayor’s office was routinely deleting emails? Policy chief Michael Kineavy said he assumed the emails were being backed up (blame the IT department!)–and described his practice of emptying his inbox every day and then emptying the deleted items folder for good measure. Kineavy defended his actions as just his way of being neat and not an intentional circumvention of the public records law, but many viewed the behavior as classic “double-deleting”–making sure there was no data available to be backed up when the nightly backups ran. The Mayor’s office quickly improvised a recovery strategy and found over 5,000 lost emails from backups of other people who had corresponded with Kineavy, but warned that a complete recovery would cost over $250,000 to do a forensic reconstruction of the hard drive.
New software and services are quickly making the whole story sound like an unsophisticated, pre-Internet tale of backwards bureaucracy. Companies like Needham-based Sonarian now offer a hosted email archiving solution–and tools to archive instant messaging and social media–that will eliminate excuses and make it reasonable (and accurate) for a staffer to say, “I assumed it was all backed up.”
These tools are also changing the business–and arguably career–model for information technology professionals. In the past, companies invested a ton of time and money in hardware and redundancy in-house to ensure business continuity and disaster recovery. The first wave of change came through software as a service (SaaS)–the idea that instead of say setting up my own perfect backup plan, I could just pay someone else to provide the service. We quickly transitioned away from swapping tapes and installing software to configuring lightweight agents on servers to push the backups out to some offsite backup farm. Personally, we discovered services like Carbonite and Mozy to ensure our laptops were safely backed up without having to think about it.
But cloud-based services take it all to the next level. At Carbonite…I suspect there is a server room somewhere in Boston with row upon row of network storage array devices where all those backups (including my own) live. In this sort of set up, a handful of engineers are responsible for building a bulletproof system and monitoring it constantly to make sure it works. The cloud approach forgoes the server room altogether in favor of Amazon Web Services. I assume there are no employees of Sonian loading tapes or swapping out failed RAID controller cards at 3am. Instead, the systems architect builds a solution of services in an environment of redundancy and scalability instead of having to create and manage that environment.
From a business model perspective, the company looking to provide a valuable service is liberated from the hardware and “spontaneous human combustion” type problems endemic to life as a systems administrator. They can focus on the service, not the exceptions and crazy, unpredictable failures.
From a career development perspective, the systems administrator’s role is changing. It has been a long time since anyone could lock themselves in an IT room and segregate from developers, but increasingly, if you see your job as performing miracles and keeping things working smoothly–your days are numbered. We have to build solutions today–and the mindset of assuming everything will fail, so we must plan around it–kills the creativity needed to architect exceptional service. Instead, we need to find systems in the form of services we can trust and develop solutions. If you are still trying to figure out how to get Microsoft SQL Server log shipping to work reliably…it’s time for you to reboot. It doesn’t work reliably enough.
Cloud computing is not magic. I’ve been hearing about SaaS and “the cloud” for years–mixed in with a bit of talk about Service Oriented Architecture (SOA). Lots of buzzwords and excuses to have industry conferences, but what does it all really mean?
- It means if you, as a technology professional, are still banging your head against the wall solving stupid problems, you need to stop. Find services you can trust and create solutions that you are proud of instead of collecting war stories about how things all worked out in the end. Let other people manage the things you cannot control instead of trying to build redundancy everywhere.
- It means that if your business is pouring money into capital expenses and extreme specialists to build a system that won’t break…you are focusing on the wrong thing. You need to build a business that can be exceptional, not one that is less likely to fail.
{ 2 comments… read them below or add one }
Excellent points to summarize with – I particularly like ” Find services you can trust and create solutions that you are proud of… “
I think this is right on target. One thing you didn’t fully cover was how to know which services (beyond email) it makes sense to move to the cloud. It seems to me that the answer is any service that is largely commoditized (i.e. your needs aren’t terribly unique) and with which you haven’t developed a unique competitive advantage. Email is a good example because it’s a mature, relatively standardized service, and very few organizations have customized their email systems for competitive advantage. (Lotus Notes/Domino applications are an obvious counterexample, but even that could be usefully moved to a hosted Domino provider.)