3 ms·

I was one of the maintainers of Flynn for a period of time and I still use it on a personal cluster that has been humming along problem free for several years.

by jpgvm 6y ago

I was one of the maintainers of Flynn for a period of time and I still use it on a personal cluster that has been humming along problem free for several years. I'm very sad to see it go of course but the writing was on the wall for a long time after the demise of all non-k8s container management systems. This is the ultimate fate of tools like Nomad and services like ECS. k8s changed the game by providing a target API/platform that gained enough traction to force each of the large providers to support a managed implementation of it. I miss my time working on Flynn, we were a very small team but we wrote some good software and I learnt a ton in the process.

mwarkentin 6y ago
I don’t see ECS going anywhere. AWS still maintains old services like SimpleDB and ECS development seems to be speeding up if anything.
- jpgvm 6y ago
  We will see. My prediction is ECS becomes pretty much a zombie service. It's neither easier to use or better than managed k8s in any dimension that matters so overtime will bleed even more market share. What AWS has to be careful with here is that when people make the move to k8s they might depart AWS altogether and go to GCP/Azure as you are throwing away a lot of their lock-in in the process. This would be less of a problem if EKS was better so they probably need to be spending more time there than ECS, that ship has sailed.
vforvendettador 6y ago
Disagree with your assessment for other container orchestrations. Agreed that k8s has changed the game, but ECS has the support of the behemoth that is AWS and it has huge existing applications relying on it. ECS Fargate makes it easy to deploy containers easily. Note that I don't work for AWS, but do use AWS, Azure and GCP cloud services, I don't have any vested interest in ECS. I do believe that ECS has a long life in its future.
- jpgvm 6y ago
  Long life isn't the same as meaningful market share. I find it much more likely that ECS as the orchestration layer but stuff built from it (Fargate) live on as backends to k8s API.

by jpgvm 6y ago

mwarkentin 6y ago

I don’t see ECS going anywhere. AWS still maintains old services like SimpleDB and ECS development seems to be speeding up if anything.

jpgvm 6y ago
We will see. My prediction is ECS becomes pretty much a zombie service. It's neither easier to use or better than managed k8s in any dimension that matters so overtime will bleed even more market share. What AWS has to be careful with here is that when people make the move to k8s they might depart AWS altogether and go to GCP/Azure as you are throwing away a lot of their lock-in in the process. This would be less of a problem if EKS was better so they probably need to be spending more time there than ECS, that ship has sailed.

vforvendettador 6y ago

Disagree with your assessment for other container orchestrations. Agreed that k8s has changed the game, but ECS has the support of the behemoth that is AWS and it has huge existing applications relying on it. ECS Fargate makes it easy to deploy containers easily. Note that I don't work for AWS, but do use AWS, Azure and GCP cloud services, I don't have any vested interest in ECS. I do believe that ECS has a long life in its future.

jpgvm 6y ago
Long life isn't the same as meaningful market share. I find it much more likely that ECS as the orchestration layer but stuff built from it (Fargate) live on as backends to k8s API.