DevOps is an exciting paradigm that’s being adopted in many IT organization to bring agility for rapidly delivering and managing required infrastructure to support continuous integration and continuous deployment (CI/CD) efforts. By applying agile principles to IT Operations management, DevOps enables the latest application features to be continuously rolled out to production.
In a DevOps environment, enterprises need a framework to monitor and manage elements across a hybrid–or distributed–IT infrastructure. Traditional IT operations tools can’t deliver the required visibility because they start with the assumptions that your infrastructure is on-premise, behind a secure firewall, and an in-house asset. These assumptions have changed with the adoption of cloud computing and enterprises are now more often using a combination of private clouds, public clouds and on-premises datacenters.
Here are some IT operations best practices for enabling DevOps:
- Single pane of glass for shared visibility – It is critical to have one source of truth about the state of IT resources that software developers and IT operations professionals utilize. Performance, availability and health metrics before and after releases should be used to enhance coordination and reduce bottlenecks.
- Responsive monitoring capabilities for all infrastructure resources – In a DevOps environment, IT elements are constantly evolving, and enterprises must come up with new ways to monitor these dynamic infrastructures. Breadth and depth to monitor physical and virtual elements throughout the IT stack and templates to accelerate the on-boarding of new elements is key for maximizing visibility and delivering rapid and reliable performance.
- Proactive alerts to expedite resolution – Enterprises adopt DevOps to bring customer-centric solutions to market faster. DevOps teams need to focus their time and energy on building and enhancing applications instead of reacting to problems (or incidents). Unified alert management capabilities make it easier to identify root cause of incidents across distributed environments for decreasing mean-time-to-detect (MTTD) and mean-time-to-resolution (MTTR).
- Application and Infrastructure performance management for rapid delivery of resources – Application updates are deployed more quickly, and patches are released frequently in a DevOps environment. IT teams need to correlate (with artificial intelligence and machine learning capabilities) application performance changes with key events to keep pace with a continuously changing environment. These changes include code changes, configuration changes, infrastructure changes that support application performance for delivering optimal performance.
IT Ops - Life Before and After DevOps
We know that the new DevOps approach is changing the way software developers and operations interact and collaborate.
What does that mean for the job of an IT Operations Manager? Let’s look at a day in the life before and after moving to a DevOps approach for an operations guy. Let’s call him Bob...
- Life Before DevOps… Bob had to be involved in every release. Dev teams had scripts they wanted, and engineers need VMs to be provisioned and/or configured.
- Life After DevOps… VM provisioning and configurations are automated through config management recipes or scripts that are embedded with each code deployment, so Bob doesn’t get pulled into the tactical fire drills for releases.
Monitoring Production Apps
- Life Before DevOps… Bob has multiple apps that are in production that he is responsible for once they go live. He is reactive to issues in the environment because they don’t appear on his radar before then.
- Life After DevOps… Because ops participated early in the process and he’s tracking the app from dev stage, Bob has insight into the potential performance issues from the outset and has been able to mitigate many of them, working with the developers.
- Life Before DevOps… New servers are being added by engineering all the time. He may or may not know when they are added, but he sure is responsible for the performance of the environments.
- Life After DevOps… Automated scaling allows for new servers to be provisioned and configured. If he has a solution like Vistara, they’re automatically discovered and managed according to the rules he’s set, whether they are in dev, test or production.
- While Bob’s day to day tasks have changed, he still needs the right tools to enable him to operate a dynamic infrastructure that supports dev, test and production environments and must be managed with a service-centric view.
About the Author
Aki Naiko leads marketing and communications at OpsRamp. Aki joined Vistara from Cisco, where she drove joint marketing with service providers and cloud partners globally.
This article first appeared on OpsRamp.com.