Resource monitoring

This page provides an overview of resource monitoring, alert configuration, and scaling recommendations for effectively managing Prophecy deployments.

Monitoring

To better understand how Prophecy is using resources, you can find an overview in Settings > Admin > Monitoring. The different components correspond to the microservices that run on the Kubernetes cluster.

When usage levels reach yellow or red zones, you can choose whether to allocate additional resources. In most cases, temporary spikes in CPU and memory usage type do not necessitate action.

Resource	Usage	Suggested action
CPU	Red zone persists for more than 1 hour	CPU allocated to the service is under-provisioned. Increase by 10%.
Memory	Red zone persists for more than 1 hour	Memory allocated to the service is under-provisioned. Increase by 10%.
Storage GBytes	Yellow zone is reached	Increase the storage capacity of the PVC in your Kubernetes deployment that corresponds to the service.
Storage File Count	Yellow zone is reached	Increase the storage capacity of the PVC in your Kubernetes deployment that corresponds to the service.

Getting help

If you need help increasing resources for a service in your Kubernetes deployment, reach out to Prophecy support.

Alerts

To set up automated alerts about your resource usage, visit the page on Alerts Configuration.

Scaling

Scaling is taken care of by the Autoscaler component in our platform. Therefore, if something happens like the number of users increases by 10 times, the Prophecy deployment can be scaled appropriately.

The following are estimated recommended cluster sizes depending on the number of concurrent users:

Number of users	25	50	100
CPUs	56 vCPUs	96 vCPUs	180 vCPUs
Memory	128 GB	256 GB	440 GB
Disk space (with backups)	360 GB	720 GB	1440 GB

info

The recommended resource may vary based on the intensity of the usage of each developer. The numbers presented above are based on the average recorded usage of Prophecy customers.