A couple of years ago, the Huffington Post ran an article called “The Courage to Screw Up: Why DIY is Good for You.” The article extolled the virtues of making things yourself, explaining that the learnings outweigh the downsides. The author found that most DIYers have one thing in common: the courage to screw up. While the idea of making something yourself sounds inspiring – and sometimes it can be downright cathartic – there are some things you simply can’t afford to screw up.
Here’s an example. Last year, I decided to build my own dining room table. I’m fairly handy, although certainly not a master carpenter, and I wanted to build something my family could be proud of (while saving a little money). With cost being a driver, I found online plans for the table, claiming it would cost only $100 to build. I set about finding some beautiful wood, the right stains and varnishes, the proper tools, online construction plans and other materials. I spent about $300 in supplies, which immediately put me $200 over budget. I then proceeded to shell out an additional $1,800 in tools on top of the $1,000 worth of tools I already owned, just to make sure I did the job right.
It took three months of construction working nights and weekends before I finished. I worked HARD on this table, and had a few black-and-blue fingernails to prove it. However, it wasn’t architected properly; the joinery isn’t ideal, and I’ll probably have to rebuild portions of it.
In the end, I could have taken the money and time I invested in my DIY project and purchased a $10,000 table that I wouldn’t have to touch for the next 20 years. But if I were given the opportunity to remake the build vs. buy decision, I would build the table again. When I see the faults in the table, I think of the time I spent with my children assembling the top. I can look back at those moments fondly. Ultimately, the faults don’t influence my bottom line. Can you say the same of your infrastructure deployments?
I learned a valuable lesson from my table-building adventure: some things are better left to the experts. Being in the always-available production services business – with Sungard Availability Services (Sungard AS), the leader in production and data resilience – I can tell you that the DIY mentality can have a very negative impact on your IT budget.
Before you DIY your cloud migration, ask yourself a few questions:
- Focus on what you are good at; are you providing real value to your customers?
- Do you have the resources to architect in the cloud?
- Do your resources have the skills to do it correctly the first time?
- Are you getting all you can out of your infrastructure?
- Will those same resources continue to support you after you deploy?
If the answer to any of these questions is “no,” it’s time to get help.
Case in point: We work with a transportation company that develops software & hardware leading to more productive, safer and ecological driving and improved customer service. The company uses some 80,000 “IoT” (Internet of Things) devices on trucks, including telemetry and GPS data, to manage its peak workload on Monday mornings (things are slower on weekends) with 2G connections.
Currently this company writes batches to their data center (executing a series of jobs in a program on a computer without manual intervention) – a process which is not dynamic. Knowing they were not equipped to build and protect their own data center, they turned to Sungard AS for an AWS (Amazon Web Services) IoT gateway to serve as a message broker, rules engine, security check and identity check, and provide device shadows and a shadow service (making automatic backup copies or snapshots of computer files or volumes).
In the first week, the company executed some 5,361,275 AWS IoT rules, on top of 2,675,437 AWS IoT messages. This is on top of some 116,675 connections from trucks, auto scaling from 2 to 9 r4.4xlarge instances – both up and down – with up to 3,000 docker containers per host.
How does it all translate into benefits for their drivers and customers?
- Real-time analytics for the entire fleet
- Updates pushed to devices at turn-up
- Auto-scaling across availability zones (AZs)
- Alerting on dynamic metrics
- Continued/improved audit compliance
Having established business priorities for resiliency with this transportation firm, along with their critical applications and dependencies, we then worked with them to configure a recovery process.
If your goal is to build a data center in the least amount of time while dealing with an event at home base, that requires a partner who knows what they’re doing, leaving you free to deal with your business processes and innovations.
In the end, what did our customer learn from working with a partner, versus doing it themselves?
- Solve for today
- Architect for tomorrow
- Focus on your business
- Deliver on fully-recoverable production services
Speaking from my own experience, I’m proud to celebrate this Thanksgiving with dinner served on the table I built with my own two hands. In other words, DIY is great for anyone on a personal level – but not always the right answer for your business. For more insights into the technologies involved with a hybrid IT implementation, please attend our Partner Session at AWS re:invent on Wednesday, November 29 from 12:50 to 1:05 p.m.