How will I be charged and billed for running my containers on PodSpace?
There is a monthly base subscription fee which covers the access to the PodSpace platform, and has some basic features included in the price. And there are extra fees that you have to pay for consuming resources like CPU, memory, and disk storage on PodSpace.
The measurement and charging of these resource costs are done in a revolutionary way unlike other public cloud providers: you really only pay for the resources that you have actually consumed. Most providers will charge the hourly cost for an instance regardless of the amount of resources that were actually consumed by that instance in that hour: e.g., if you have an instant with 4 vCPUs, you will pay the hourly cost of those 4 vCPUs even during periods of low consumption, when the instance was doing nothing and did not use the CPUs at all.
With PodSpace, this is different: we constantly measure the amount of resources used by your containers, and charge you only for the actually consumed resources. If your application was idle for an hour, and consumed only 2% of a single vCPU, you will only pay the price for 0.02 CPU hours for that hour.
The same thing holds with memory too: if your application’s memory usage was 256 MB in one hour, and 287MB in the next hour, you will be charged 0.25 GB hours for the first hour, and 0.28 GB hours for the next.
How long is the billing cycle, and when does the billing occur
The billing cycle starts on the 1st day of the month, and ends on the last day of the month. During this period, resource usage charges are collected on every Monday morning to an OPEN invoice which you can view on the admin console.
At the end of the billing cycle, the open invoice gets closed and becomes a legally binding invoice which you must pay.
We strongly recommend you to monitor the resource usage of your applications constantly, and review your open invoice weekly so you can intervene in time in case one of your applications behaves erraticly and consumes more resources than it should.
How can I pay less?
There are a couple of things to consider to optimize your costs:
You have the possibility to easily scale your applications to multiple copies. Of course, doing this also means that you will have to pay more. If you currently have applications scaled to more than 1 instances, you should consider configuring an autoscaler, which will scale your application up and down depending on the actual load it receives.
When configuring your deployments, you have the possibility to specify resource requests for your applications. A resource request means that the requested amount of resource (e.g. 200 millicores of CPU) will be reserved for your application. However, this also implies that you will be billed for the reserved amount for the resource even if your application used less than that.
You should consider lowering your deployments' resource request settings (under Actions / Edit Resource Limits on the deployment page).
Optimise your app
You pay for the resources that your application consumes. So try lowering the memory and CPU footprint of your applications, and your bills will become smaller accordingly.
Which features can I use for free?
Every subscription plan includes some services that are included in the base price and are not subject to resource usage charges:
Any number of docker repositories (limited to the amount of storage space specified in the subscription plan)
Load balancers: Route and Service definitions (limited to the number specified in the subscription plan)
Automatic scaling of deployments
Application health monitoring
Multiple projects (limited to the number specified in the subscription plan)