Infrastructure automation is far from being simple. It involves integration of many types of actions, on many types of devices such as routers, servers, storage, switches, load balancers, laptops, and desktops.
However, there are several tools and methodologies available today to service infrastructure, including software distribution, server provisioning, network management automation, virtualization, virtualization management software, and storage configuration and management.
In contrast, no platform exists today to help with application operational tasks – also known as Application Service Automation. These tasks include complex application deployment, application configuration changes, troubleshooting and recovery.
There are three main areas that make application service automation different from infrastructure service automation:
Task Complexity
Many of the infrastructure automation tasks involve repetitive tasks on many devices, usually on same-server types or devices, or according to domains. The nature of automation in this case is perform the tasks on a group of servers together, or according to domains.
In contrast, application services usually involve multiple server types/domains. The processes needed to service applications are complicated, involving numerous steps, strict procedures, and a variety of machines. Anyone who has gotten the “checklist” for an application release process knows how complicated this can be, and how unlikely it is for two people to actually do every step in exactly the same way, regardless of how well documented the process is.
Dependencies
With infrastructure service automation, the focus is on updating the device. With application service automation, teams must understand the application architecture – which servers are mapped to each other and how all servers depend on each other. In order to automate a process, they need to understand which particular application servers connect to precisely which database servers.
Order
When servicing infrastructure, it doesn’t matter if this router gets its new configuration first or which server gets its security patch first. But for application service automation, order of operations matters. To update an application, the application logic is important. You might need to take down the database server first, then the application server, and finally the Web server, before making changes. And if you do not have maintenance windows, you may need to bring down a very particular part of the application first, make your changes, and only after that part is back up and running, do the next group of servers in a rolling update.
Today’s application service teams require automated tools that understand the architecture and logic inherent in the applications they service. To learn more about how Nolio supports task complexity, grouping, dependencies and order with Nolio Automation Center, please visit www.noliosoft.com.
application service automation