Virtualization Technology News and Information
iQuate 2017 Predictions: Will the fallacies of distributed computing hold true in 2017?

VMblog Predictions 2017

Virtualization and Cloud executives share their predictions for 2017.  Read them in this 9th annual series exclusive.

Contributed by Paul Mansfield, CTO of iQuate

Will the fallacies of distributed computing hold true in 2017?

22 years ago Peter Deutsch crafted the 7 assumptions that developers and architects tend to make and which will prove wrong in the long run. James Gosling (the creator of Java) added the 8th in 1997.


1. The network is reliable

The assumption that nothing ever goes wrong with the network is probably even more detrimental. Software and application architecture must take network failure into account and the information on network /service dependencies must be available at our finger tips. The knowledge of integration points and the service calls made at these points will continue to be important.

2. Latency is zero

Developers and architects need to work on ways to reduce the impact of latency by optimizing the number of remote calls, eliminate queueing. Having a complete view of service dependency maps including all new points of interaction such as IoT devices, and tracking of latency times between calls will also need to be top of mind. Implementation of patterns such as circuit breakers, bulk heads help in introducing robustness in the application

3. Bandwidth is infinite

The rate at which our bandwidth has been improving is impressive but so have our consumption habits. Users and applications are increasingly hungry for data. The implication at the technical level is ensuring that communication is optimized i.e. the message sizes are not too big and work under the assumption that data-loss can occur.

4. The network is secure

Humans are wired to break into things and steal. The most recent high profile event was Denial of Service (DDOS) attack that originated by inundating the Dyn servers. If you recall, Dyn provides Directory Naming Service. Data could (knowingly and unknowingly) sit anywhere in the world across multiple clouds, nefarious services such as Ransomware as a Service we have an environment that is ripe further cyber break-ins.. It will be crucial to design and architect solutions with security in mind.

5. Topology does not change

Change is an inherent reality of the network. With the usage of services across multiple cloud providers, distributed services, containers, mobile and IoT devices coming in and out of the network, the network not only is changing but it is changing at faster pace than ever.

6. There is one administrator

If you ask any developer, architect if the service is managed by ONE person, the answer will unanimously be a NO. It is managed by a team. But un-fortunately the design and development often assumes that there is one administrator.

The design and development should assume multiple teams with different domains of expertise and ignorance. This gets exacerbated with micro-service architectures, multi-cloud deployments, multi -application architecture and diversity in connected devices.

7. Transport cost is zero

We already covered that bandwidth is not infinite and closely tied to it is the fact that bandwidth is not for free. Data centers of today are not bound by four walls. With the increase in adoption of external cloud providers, architects and developers must consider inbound, out-bound, and content delivery costs.

8. The network is homogenous

This might seem like the most obvious one, that of course the network is not homogenous that there are different types devices, applications, mobile, devices etc that are part of the application. This becomes very interesting when you combine with other assumptions around latency, bandwidth, changes etc. Now the developers, architects and operations must design and manage with heterogeneity in mind and remember that each component behaves differently.


About the Author

As Chief Technology Officer, Paul Mansfield is responsible for the company's technical vision and strategy with a focus on delivering product offerings which are innovative, best in class and, above all, client focused. Prior to iQuate, Paul held senior technology and management roles at Logica, S1 Corporation, Allianz and Macquarie Group across Europe and Asia Pacific, where he gained a reputation for outstanding technology leadership and strategy.  


Published Thursday, December 29, 2016 9:01 AM by David Marshall
There are no comments for this post.
To post a comment, you must be a registered user. Registration is free and easy! Sign up now!
<December 2016>