I don't understand how machine learning fits into literally just an option to power down a machine on a schedule?
I know when my business does not need a particular environment up because it's my business... why would I need something to train that and than apply it, probably incorrectly, for me?
Regular reminder for readers: If a service exists that claims to meet a need but you do not have that need, it means you are not a customer for that service, not that the need does not exist. If the need is not large enough, or if not enough customers exist that have that need, then the service may fail as a business, but that does not mean that there was no need at all.
If you "know when my business does not need a particular environment up" then you are not a great customer, but other people might be.
Did you look at the product? It gives you schedule suggestions. So you have to know your schedule to pick the right one, which means you could use something that allows you to build a schedule.
If the product is for people who think adding unnecessary machine learning to simple tasks then sure, I'm not the audience but a condescending "you don't get it" is not really adding any clarity. Do you get it?
I didn't say "you don't get it", I said "you are not a customer". Not having the same needs as someone else does not imply that you do not understand those needs or are unable to reflect on whether you might share those needs. You know your needs quite well and I assume you considered whether the product met a need of yours.
But just saying "I don't need this and it probably wouldn't work even if I did" doesn't add any clarity either. :-)
What would add clarity is to ask the creator whether they see it being applicable in a particular use case, and you can give your specific use case, and give them the chance to say "yes, here's why" or "no, the product doesn't help you". Then future readers can see the use case and decide if they have similar needs to you and it would greatly help their purchasing decision.
Honestly, I wish more product discussions were focused like that. "Here's my situation" and the entrepreneur gave an honest "yes, here's why" or "no, the product doesn't help you". But it typically is just a commentator saying "I don't need this because I solved that problem myself" and the product person claiming that it works in every use case, so other readers don't get as much of a benefit from the discussion.
Software only provides one recommendation (confidence level of recommendation can be adjusted). Customers can confirm the recommendation by visually inspecting the context sensitive data graphs and let the software execute the schedule. In many large environments often customers really don't have much of an idea as to when their long running instances are actually being used. This feature provides great insight into instance utilization and possibility of significant cost savings by automating schedules. Agree it may not be relevant/appplicable in every environment.
This seems interesting and sounds like a fun project - but... How many EC2 instances can just be turned off and on? If they can, they're probably part of an autoscaling group and already scale based on required capacity... Or they're workstations and can be turned off on a schedule (as mentioned in the article), ie: work hours only.
As much as I want to, I can't think of a purpose for this. Would be fun as a Kubernetes scheduler! :D
> How many EC2 instances can just be turned off and on?
I could actually imagine situations where you have loads that also follow a schedule -- for example, if you have a ton of data and customers can schedule summary generation/analyses, a lot of them will tend to schedule them on the first of the month (that is, after the last month has ended).
In a case like that with a predictable load, you could boot up a bunch of instances a bit before the end of the last day of a month, then turn them off (to go back to normal loads) a bit into the second day of the new month.
Machine learning could help you pick out other peaks, like say, every Monday or Sunday.
> In a case like that with a predictable load, you could boot up a bunch of instances a bit before the end of the last day of a month, then turn them off (to go back to normal loads) a bit into the second day of the new month.
If it's predictable, no need for machine learning at all.
> "Machine learning could help you pick out other peaks, like say, every Monday or Sunday."
In our view combination of machine learning and elastic provisioning in public clouds is incredibly powerful and will be fundamental to resource provisioning in public clouds. As environments become large/complex/dynamic, there will be no easy way to monitor provisioning vs utilization and adjust them. Machine learning driven resource provisioning will become a necessity. This is only one example of such possibilities.
I don't know how to setup an autoscaler group on EC2. But if I used this wouldn't I not have to set one up? What if you don't have perfect information about your fleet and some instances are workstations?
This is a rather low content/high marketing post, but the idea is interesting. I'd be more interested in a description of how it actually works, but I guess that's FittedCloud's secret sauce.
