SRE Tooling and Processes at HCL DevOps [excerpt]

The maturation and adoption of SRE standards by the market have made SRE the new DevOps 2.0 in the eyes of many organizations.

While SRE evolved from tried and tested DevOps practices initially laid out and popularized by Google, in reality, most companies will have different SRE philosophies and principles custom-tailored around their own business needs.

Let’s look at the best way to leverage DevOps tooling, what processes and philosophies organizations use to take their DevOps program to the next level, and how to get the most out of HCL’s DevOps offering.

DevOps is a set of practices used to build a collaborative culture between the development and operation teams that increases productivity while decreasing time to market and cost for your business.

Let’s take a look at a couple of SRE standards and how your company can benefit from them:

1. Reduce organizational silos

Whether using a process that necessitates collaboration between development and operation teams or having both teams sit in the same room physically or virtually creates shared ownership of the path your product takes through each phase up to the end-users. The collaboration and discussions among development, security, marketing, and other relevant teams are the primers in ensuring that planned features and changes can work in the real world.

2. Accept failure as normal

SLAs, SLOs, and SLIs are a must when measuring performance and critical for the business’s success. Still, errors will occur, and when they do, having a well-prepared rollback plan will prevent spending time troubleshooting bugs. Accepting failure as normal increases the collaboration and focuses the teams on better preparing for errors.

3. Implement gradual change

Breaking down big changes into smaller releases eases the job of both engineering and operations teams. It reduces the cost of failure by having less redundant work for operations and decreased investigation time for developers.

4. Leveraging tooling and automation

Automation is key in creating long-term value and will start with identifying what should need automation from the collaborating teams’ outputs. As a rule of thumb, automation should be used for any repeating task as long as the time spent building the automation is considerably smaller than executing the task itself.

However, a low level of toil is acceptable and encouraged, and most companies aim for somewhere between 10 and 20 percent. HCL Accelerate can help tremendously to overview the sticking points that occur most often and identify the most critical steps to improve first.

5. Measure everything

Too much data can divert from the real problem, and too little will trigger stagnation in your process, but we need to look at the differences between logging, measuring, and alerting. The balance will differ from case to case and industry to industry, and SLAs, SLOs, and SLIs levels will be the main topic for the development, operations, sales, and other teams. HCL Accelerate, HCL One Test, and HCL AppScan integrations give a clear vision for your product cycle’s usability and security perspective.