This questionnaire is aimed to help you/us find the most suitable way of deploying Sysdig Secure for Cloud.
Despite wanting only to trial it, we suggest that you deploy, and test it, in th most-similar situation to what you have on your production environment.
We are aware that current examples don't suit all situations, and we will keep improving them to be as configurable as possible. Contact us with these questions answered to help us.
Sysdig Secure for Cloud is served in Terraform AWS, GCP and Azure modules, and we also offer AWS Cloudformation templates
- does your company work under an organization (AWS/GCP) or tenant (Azure)?
- if so, how many member accounts (aws) /projects (gcp) /subscriptions (azure) does it have?
- regarding of the number, how many accounts would be required to enroll in the secure for cloud setup?
- do you have dynamic accounts/projects/subscriptions? what's their lifecycle?
- does it have any landing such as aws control-tower?
- if so, how many member accounts (aws) /projects (gcp) /subscriptions (azure) does it have?
- sysdig secure for cloud is presented in different compute workload flavors; ecs on aws, cloudrun on gcp or azure container instances on azure, plus a K8s deployment an all three clouds, plus apprunner on aws (less resource-demaing than ecs, but region limited)
- in case of ECS or K8S, do you have an existing cluster you would like to re-use?
- (aws-only) do you have existing aws cloudtrail?
- is it an organizational cloudtrail?
- does the cloudtrail report to an SNS? if no, could you enable it? ingestor-type:
aws-cloudtrail-sns-sqs
- is the S3 bucket of that cloudtrail in the management account or a sepparated member account?
- does the cloudtrail report to an SNS? if no, could you enable it? ingestor-type:
- if it's not organizational, does each trail report to the same s3 bucket?
- if so, does that S3 bucket already have any "Event Notification System"? Is it an SNS we could subscribe to? ingestor-type:
aws-cloudtrail-s3-sns-sqs
- if so, does that S3 bucket already have an "Amazon EventBridge" system activated? ingestor-type:
aws-cloudtrail-s3-sns-sqs-eventbridge
- if so, does that S3 bucket already have any "Event Notification System"? Is it an SNS we could subscribe to? ingestor-type:
- whether it's organizational or not, could you give us a quick picture of the account setup in terms of purpose?
- is it an organizational cloudtrail?
- how many regions do you work with?
- is secure for cloud to be deployed on the same region as your existing resources?
- if not, explain us your current region setup
- (aws-only) if in previous point you said you have a cloudtrail, cloudtrail-sns, or cloudtrail-s3, in which region is it?
- how do you handle IAM permissions? would you let our Terraform scripts set them up for you, or you want to set them yourself manually? any restriction we may be aware of?
- how do you handle outbound newtwork connection securization? does your infrastructure have any customized VPC/firewally setup?
- Deployment type
- are you familiar with the installation stack? Terraform, Cloudformation, AWS CDK, ...? Do you use any other InfraAsCode frameworks?
- if you want to use Kubernetes compute for Sysdig deployment, what's your current way of deploying helm charts?
In what Sysdig For Cloud Features are you interested in?
- Runtime Threat Detection
- Compliance (cis benchmarks and others)
- Identity and Access Management
- Scanning
- Push-based registry/repository image scanning
- Runtime workload image scanning (ecs on Aws, cloudrun on GCP, or container instances on Azure)
- Note: Sysdig offers many other ways of performing scanning, and we recommend you to Check all Scanning options in the Vulnerability Management to push this task as far to the left as possible (dev side)
Single single- |
Organizational organizational- |
|
---|---|---|
Deployment Type | all Sysdig resources will be deployed within the selected account | most Sysdig resources will be deployed within the selected account, but some require to be deployed on member-accounts (for Compliance and Image Scanning) and one role is needed on the management account for cloudtrail event access |
Target | will only analyse current account | handles all accounts (managed and member) |
Drawbacks | cannot re-use another account Cloudtrail data (unless its deployed on the same account where the sns/s3 bucket is) | for scanning, a per-member-account access role is required |
With both examples single
and org
, you can customize the desired features to de deployed with the deploy_*
input vars to avoid deploying more than wanted.
Cloud | Example Options |
---|---|
AWS | K8S -k8s , ECS -ecs , AppRunner -apprunner |
GCP | K8S -k8s , CloudRun |
Azure | K8S -k8s , AzureContainerInstances |
We enable following optionals, to allow user to re-use their pre-existing / configured resources.
Cloud | Optionals | Related Input Vars | Other |
---|---|---|---|
AWS | Cloudtrail | single: cloudtrail_sns_arn organizational: existing_cloudtrail_config |
- |
ECS, VPC, Subnet | ecs_cluster_name , ecs_vpc_id , ecs_vpc_subnets_private_ids |
if used, the three are mandatory | |
GCP | - | - | - |
Azure | ResourceGroup | resource_group_name |
- |
ACR | registry_name , registry_resource_group_name |
- | |
* | Compute Workload | - | All clouds allow Sysdig Secure for cloud to be deployed on a pre-existing K8S cluster |