FAQ
Last updated
Last updated
It usually takes 3 seconds to launch a new pod in Pi. However, there are several factors that could affect the performance, e.g. the number of containers in the pod, image size, application startup speed, etc.
No, you don't need to. Pi is a serverless container platform, which means that you can just run containers without worrying about the underlying infrastructure. Things like capacity planning, scheduling , resource utilization and cluster autoscaling are all gone.
By default, you are limited to running up to 100 pods per region, though you may .
By default, Pi creates a private network dedicated for each account, so as to separate different customers:
When you create a new pod, it is placed in your own network.
Every pod receives a private IP address (specific to the network it resides in). The IP address is attached exclusively with the pod and is only returned when the pod is terminated.
A pod is accessible in the network where it resides (using its private IP address), but no other networks.
All pods come with the access to the Internet, however none is by default accessible from the Internet. This is adequate for most non-public-facing pods. For public-facing pods, and are demanded.
Service serves as load balancer in front of backend pods. For more detail see the background on .
The data stored on the pod's
rootfswill persist only as long as that pod exists.The data that is stored on additional volumes will persist independently of pods.
Cloud
Region
Location
Zones
Google Cloud
gcp-us-central1
Iowa
gcp-us-central1-a,gcp-us-central1-c
Except as otherwise noted, our prices are exclusive of applicable taxes and duties, including VAT and applicable sales tax.
Use command:
Similar with other platforms like : Pi does not run in our own datacenter, it runs on . Therefore, the region and zone naming in Pi follows the pattern of GCP: