Network, Regions and Zones
Cloud infrastructure setup starts with creation of VPCs, subnets, NAT gateways, availability zones and
routes. Automation of VPN and hybrid connectivity calls in this layer of automation. For compliance with
regulatory standards like PCI-DSS, it is required that production and non-production environments are in
a different VPC. Multi region deployments are a common scenario.
With the network infrastructure in place, next comes storage, compute and other application services
like databases, object store and Encryption keys. These need to be segregated as per application
boundaries with appropriate security groups and IAM/AD policies. User access controls like
Just-in-time access, single sign-on and password management are performed in this layer.
Backup-and-restore configuration and disaster recovery automation is done here.
Application Provisioning and Operations
With the advent of microservices and Docker, infrastructure is fragmented requiring a lot more advanced
automation. In this layer of the stack, DuploCloud manages Kuberenetes clusters, serverless deployments
and ETL pipeline automation that includes Spark, Athena, Jupyter etc. The load balancer, auto-scaling,
DNS automation, health checks, rolling upgrades, Blue Green deploys and Disaster recovery are some of
the other key functions performed here.
Engineering teams spend a majority of their time in debugging. While there are tons of tools like ELK,
Sumologic, signalfx, Cloudwatch, Sentry that implement these functions, they are still required to be
orchestrated and configured in the context of the application. DuploCloud performs this orchestration
and creates per application dashboards with all the relevant infrastructure metrics. We have out-of-box
integrations with most logging, APM and alerting tools..
Unified CI and CD
DuploCloud has a built-in CI/CD saving the manual effort to set up build systems like Jenkins. With
DuploCloud we treat builds as transient application components that execute side-by-side in the same
topology as the application itself. Each build triggers a set of docker containers that run the spare
capacity of the application servers using the same automation as the application provisioning itself.
The containers are cleaned up after the build is over. This enables Unification of CI and CD.
The user is still not mandated to use the DuploCloud CI/CD. They can continue to use their current
system and integrate with DuploCloud’s provisioning layer via APIs.