CIO in Cloud. The Worst Job in IT.
First of all, I have to admit that if I was a CTO of a cloud provider, I would be totally unhappy. Technical staff in Cloud industry have much harder life than anywhere else!
Listening to endless speeches about innovation from Top-managers and ordinary Sales of Cloud-providers I was surprised to learn that technical directors of such companies are super conservative and they hate to be the first to try something new. Most of new technologies (SDN, virtual storages, etc.) came to providers’ datacenters with long delays after they were broadly available on the market.
What has made Cloud CTOs so cautious? What has happened to their courage? The answer is in their hard job.
From Tech. Support point of view, infrastructure of Cloud is like a bus full of children. They constantly change their seats, jump, blow up firecrackers, throw balls at the driver. And to avoid a road incident a transport company has to implement expensive things such as triple glasses, two drivers, and hang sensors everywhere in the cabin. In Cloud the Technical Support team never knows in which second one of the customers suddenly increases traffic in 10 times.
On the other hand, all technical errors in cloud are much more transparent. It is a hard task for management of any organization to control its CIO. In most cases, only a dynamic of IT-cost is a subject of regular tracking. And maybe not-working mail of CEO is also an indicator that their IT Dep is not perfect. But in Cloud business everything is quite glassy. The customers have a detailed SLA in their contracts (near ‘Penalties’ section) and if something goes wrong, customers complain directly to provider’s top-management. Unfortunately, no chance to hide.
Technical staff of Cloud providers are permanently squeezed by meeting SLA and cost reductions. So, the first thought of CIO’s mind about a new technology in Cloud is how this innovation could drop their infrastructure.
If you can standardize your stuff, you can compare them with others. If you can compare, you can learn where you are losing. If you learn, you can optimize. It is hard to find which one from two paintings is better, but you can easily do it for two TV-sets. Cloud is all about standardization. If one customer needs CRM and another one needs Mail, nobody will make a cloud service for them. But if we have 100 customers, who need Mail we can create one Mail service and subscribe them all. However, if your product is standardized, a customer can easily compare it with a similar offer of your competitor. That is why competition on the cloud market is very high. Cloud providers have to keep optimizing their core services permanently. Cutting cost and increasing SLA is exactly what Cloud Tech. Teams do most of their days.
Many CIO wants to transform their IT-departments into a services-oriented organization, but Cloud providers cannot afford not to do it. Having formalized detailed obligations to customers (SLA), providers have to use internal SLA on each level and track them carefully. Customers through compensations can control how things are going, and this monitoring system is much crueler than any SCOM or PRTG.
It is not a big deal, if you drive into a pothole, but if you drive in circles, you will face this pothole many time and you need to change your route immediately after the first one was found. Operations in cloud is like a circle, the similar processes for many customers that retry many times. And if a pothole appeared, it should be plugged much faster than in if you drive ahead.
Repeatable nature of operations in Cloud and strong surveillance from customers press cloud providers to optimize their operations much stronger than in an average IT team. That specific can be depressing for technicians making them factory workers on conveyor but it is good for their customers making service more reliable and cheaper.
The better Cloud is performing, the more there is routing work for Support Team. And, you know, many engineers don’t want to go to Cloud Team because of it. But, thank God, technical incidents happen from time to time. In that period, it is better not to be in Cloud Department corner of the office. How many calls management of providers have in first 30 minutes of serious technical incident? Correct, it is equal to number of customers. If customers use standard service, an error affects to all of them. And who will be blamed afterword? Of course, the technicians. Who else?!