The Future is Not Stable

While the server and network team scream for justice, the DevOps team roll out new code faster than my boss can run to the cafeteria queue at lunch time (quite fast).

New features, new features, new features!

We want new features and all the blinking lights, right now.

If the new roll-out breaks the application, we can always roll out a new fix.

There is no time for testing. What else is automatic updates for.

Just look at your mobile phone.

The frequency of applications updates are often multiple times a week.

These are not security updates. These are mostly feature updates.

Having things in the cloud made speed of change faster and easier, but as a company we still rely on our systems “on the ground”.

When we combine these systems with cloud solutions, we end up with a hybrid and moving hamburger on a dinner plate that is trying to stay still.

There is no stopping the future. Change is constant. It will just happen faster.

So why do the server and network guys scream for justice?

Well, SLA’s and demands for stability are some reasons.

How can we make something stable if the demands are to constantly change stuff?

The future will bring faster and faster change and the foundation that the applications run on can not be a static and therefor stable setup.

The foundation will need to be able to change at the speed of the application changes.

If we want to or not, we are slowly accepting that applications have problems due to change.

We know that if an application crash, we can just wait for the next update where it will probably be fixed. It is just code and code have bugs and we accept it.

But often, these changes require that the underlying infrastructure adapt to be able to get things working as intended.

Doing changes to the infrastructure on the fly has never been a good thing in the long run.

It tends to get very messy after a while.

We know. We have tried it.

So how can we do this?

Will changes to the infrastructure be accepted in the same way eventually?

Even if these changes break the infrastructure?

Will people just sit back and relax and wait for the fix?

What then about SLA’s?

I think, on behalf of the server and network teams, that we need to adapt whether we want to or not.

There is no stopping the future. Change is constant. It will just happen faster.

All links in the chain needs to be able to change instantly and we must live with messy and unstable. There is no easy way to get around it. We can’t have both worlds of fast change and always up.

“To keep up we need to mess up and flow with the changes.”

J. Lanesskog, Marin IT