This guest post was written and published by Rob Mason, Core member and multi-time successful Entrepreneur following a recent Cloud Infrastructure Roundtable at Underscore’s Core Summit.
Recently I had the pleasure of co-moderating the Cloud Infrastructure Roundtable with Ash Ashutosh at Underscore’s Core Summit in Boston. We had a great group of about two-dozen people with a wide range of experiences with Cloud Infrastructure.
About 40% of the group was working on making new pieces of Cloud Infrastructure. The other 60% were using Cloud Infrastructure for IT, Dev Ops, etc. Often, as developers and users, we get caught up in the day-to-day activities and have little time to compare notes with other experts in our areas and it was a great opportunity to gather some collective wisdom.
Here some of the key thoughts from the group:
- The software industry as a whole is shrinking as applications move to the cloud.
- Infrastructure that bridges to the cloud is still in demand, but the group saw that declining in the future as more and more infrastructure finds a home natively in the cloud.
- Most people agreed that the cloud could do things better than they could do themselves in house in terms of reliability, availability, performance, etc.
- Cloud Governance is a growing concern and more solutions are needed in this area.
- When we discussed why people were moving to the cloud, these were the top reasons:
- Cloud Initiative by company (these can often be a waste of time for vendors as there can be little commitment to them)
- Resume building by those involved, similar to cloud initiatives.
- The goal of owning less and less infrastructure over time.
- The speed at which new infrastructure can be created/brought online.
- The breath of services that are readily available in the cloud
- Cloud economics (we discussed a price benefit anywhere from 3:1 to 5:1) (important to look at the full TCO here).
- Some applications/environments cannot be moved to the cloud and there will be a long-term need around hybrid/mixed cloud and datacenter solutions for those specific areas. Identifying those areas early on is key to success.
- Switching costs (between providers) are high. Some technologies, such as Cloud Foundry, can help but are not widely adopted and do not address all the switching needs.
- Vendor lock-in was a heated discussion with people on two sides of the topic:
- Use lowest common denominator features from the vendors to make switching easier and/or wrap all vendor interfaces with their own.
- Don’t worry about it, let solutions be a mix of multiple providers – things are moving too quickly and applications are getting created and deployed too rapidly to spend time shifting solutions between providers.
Many of the topics above were only touched upon briefly and many would love the opportunity to have follow-up discussions with a deeper dive into the various areas, as will be held in future Core meetings.
It’s clear that Cloud Infrastructure continues to be a hot topic and growing space — more than 10% of the people in the discussion already had Cloud Infrastructure expenses in excess of $1M per year. Everyone in the group was very interested and active in the space and appreciated the chance to join in on the roundtable discussion hosted by our friends at Underscore.