You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Is your feature request related to a problem? Please describe.
I saw ECS, k8s and Azure Container Instance have a configuration option to set memory and cpu. As I deploy Mage on GCP, I found an issue that the cloud run job memory is not enough to run my job. Right now, the Cloud Run Jobs CPU and memory configuration will copy the config of the Cloud Run service if mage is deployed on Cloud Run service. So to increase the resource allocation, we need to set the service resource allocation higher, while it might not be necessary. Also each block might need different resource allocation.
Describe the solution you'd like
Make the configuration option to set memory and cpu for cloud run executor
Describe alternatives you've considered
Another option is to set a generic config on cloud run as a mapping to be passed to cloud run API function. This way, if GCP add new feature on their API, we can just pass the keyword argument to the metadata.yaml and the executor will pass it to GCP function.
Is your feature request related to a problem? Please describe.
I saw ECS, k8s and Azure Container Instance have a configuration option to set memory and cpu. As I deploy Mage on GCP, I found an issue that the cloud run job memory is not enough to run my job. Right now, the Cloud Run Jobs CPU and memory configuration will copy the config of the Cloud Run service if mage is deployed on Cloud Run service. So to increase the resource allocation, we need to set the service resource allocation higher, while it might not be necessary. Also each block might need different resource allocation.
Describe the solution you'd like
Make the configuration option to set memory and cpu for cloud run executor
Describe alternatives you've considered
Another option is to set a generic config on cloud run as a mapping to be passed to cloud run API function. This way, if GCP add new feature on their API, we can just pass the keyword argument to the
metadata.yaml
and the executor will pass it to GCP function.Additional context
https://docs.mage.ai/production/configuring-production-settings/compute-resource#azure-container-instance-executor
The text was updated successfully, but these errors were encountered: