Serverless Does Not Mean Brainless Part 2


Posted on April 26, 2018 by Yann Tromeur

Previous                                                                                                             Next

In yesterday’s Blog, I described a short tutorial from AWS exposing basis of serverless architecture.

Here is the second part of that article.

Things are going to change…

faster than you think

You know virtualization, cloud systems? Those disrupted (gigantically) our job as network administrator.

But developer / services providers still need a couple of guys handling things for them. Installing hardware, systems, managing them…etc.

Like fossil fuels, this is all soon gone and replaced for the good of humanity (IT people) by a better source.

I do not like this terminology, “Serverless”, because, unless mistaken, we are still cramming buildings and buildings with servers, switches, firewalls and cables.

That is not ”Serverless”, and it is certainly not sustainable energetically. Serverless does not mean less impact.

But, I guess, we are to reduce this amount of energy needed with micro services running on larger servers and Amazon and Google promise to target 100% renewable energy for their clouds infrastructure, don’t they?

Looks like Google got that target first, but the most important there is the goal not the race. And as GCP (est. 2.5 million physical servers) has 4% of Public Cloud and AWS 47 % (!) (est. 2 million physical servers).

Though the cloud footprint is not null, we are to consider two things:

  1. Sources of energy are more and more diverse, hope for a better future is here
  2. Sadly, we will need more and energy as cloud is hosting even less energy-friendly techs such as A.I. and Blockchain.

Sustainability my friend will prevail but it will be a difficult journey without impacting too much our economy.

As mentioned in last post, I gave it a go (to this AWS tutorial) and finding that Serverless will disrupt all administrator things again, made me smile again hoping for sustainability and smartness for IT.

So, what is it exactly?

Dear Wiki, thank you :

Serverless computing is a cloud computing execution model in which the cloud provider dynamically manages the allocation of machine resources. It is a form of utility computing.

Dear Forbes, thank you even more for this one line definition:

Serverless Computing has the least number of buttons to push, knobs to turn, and sliders to drag.

I guess that makes it pretty clear no?

Serverless is effortless.

Serverlessputs more time to yourself for creativity. Then, we are to be looser probably, putting more attention to details and users expectations, as we would spend less time setting things up.

Developers won’t be cranky anymore, users won’t be waiting for their turn for support, database administrators will trust the systems as those administrators (from Hell) are (finally declared useless) less occupied…

Time for ourselves to emphasize on process, security, documentation and support.


See you for part 3 !