Deployment Planning

Overview

Observo deploys its services using Kubernetes (even for VM based deployments), allowing for flexible and scalable management of resources. Customers have the option to utilize a managed Kubernetes cluster or simply provision the recommended instances based on the guidelines provided below.

Sizing

In order to determine the appropriate number of instances needed for the Observo Site deployment, the following guidelines will assist you in making that calculation. Observo recommends selecting instances with a minimum of 4 vCPUs and 8 GB of memory. We recommend that the instance should have at least twice the amount of memory compared to the number of vCPUs. For example, if you choose an instance with x vCPUs, the memory should be 2x GB. Please follow the steps outlined below to obtain the recommended number of instances, which will be one of the inputs during the creation of Observo Site.

1. Choose the instance type and note the vCPU and memory specifications of the selected instance.

2. Estimate the volume of logs generated per day, which is the log volume you plan to optimize on Observo. A good starting point would be to look at volume of logs ingested by your downstream destinations.

3. Calculate the maximum number of instances using the following formula:

Max Number of Instances = Ceil(Log Volume per day (GB) / (850 * vCPU/Instance))

By following these steps, you will be able to determine the recommended number of instances required for your Observo Site, based on the log volume and the specifications of the selected AWS instance type. This information will be crucial during the deployment of your Observo Site. Access this link to view the AWS instance types and their corresponding hourly prices.

Estimated Number of Instances

Note: Calculations assume a node type with 4 vCPUs.

Volume (GB per day)	Estimated Number of Instances
1000	1
2000	1
5000	2
10000	3
25000	8

Compute Cost

Static Cost

The static cost includes the cost of running Observo Site services to run with minimal configuration. Observo agent and other required services to manage the site and communicate with the Observo SaaS will require ~1 VCPUs and ~2GB.

Dynamic Cost

The dynamic cost is determined by the amount of data you are going to process through observo pipeline. A node type of c6g.xlarge (4 VCPUs, 8 GB) is capable of processing up to 3.4 TB of log data per day. As data-plane is auto-scalable user can set the maximum number of pods for data-plane to be cost optimised with respect to the amount of data processing.

Please keep in mind that the dynamic cost may vary based on your specific requirements, such as the amount of data processing. It's recommended to evaluate your data processing needs and adjust the number of data-plane max replicas accordingly to optimize costs.

By considering both the static and dynamic costs, you can estimate the average compute cost of running an Observo Site in your environment. It's important to review your specific requirements and consult with the Observo team to obtain accurate cost estimates tailored to your use case.\

POC Sizing Recommendations

Single Node Deployment

For a single node deployment, you must provision a virtual machine with the following minimum requirements:

• 8 vCPUs

• 16 GB Memory

• 100GB disk volume

Multi-Node Deployment

For a multi-node deployment, you must provision a Kubernetes cluster with the following minimum requirements:

• 3 nodes

• 12 vCPUs total

• 24 GB Memory total

• 100GB disk volume for each Node