Customer-Facing Reliability
Modern SREs use Nobl9
Latest News
Type Safe Validation in Go with govy
Nov 6, 2024
Alerting Center: A New Way To View SLO Alerts
Oct 23, 2024
Choosing the Right Path for SLO Adoption | Webinar with Experts
Oct 22, 2024
Reliability is Mission Critical
Reliability is More Than Just Outages
We all know the devastating impact of outages - the loss of revenue, the hit to brand image, the churn, the PR nightmare, and the all-hands scrambling that backburners projects and pushes back future revenue streams. But reliability is more than just ensuring your application is available as often as possible - it’s also about ensuring that your application performs reliably on a daily basis.
In an environment where switching costs are negligible, customers have a low threshold of tolerance for underperforming experiences. For every outage, there are countless examples of poor, frustrating performance that go unseen by organizations. These micro-outages - sometimes affecting a small segment of users for a brief period of time, sometimes affecting just one user - are massive, hidden issues that prevent revenue-driving interactions and create churn.
Nobl9, with our SLO-centric approach to reliability, brings visibility to these occurrences, empowering product teams to quickly identify and bring attention to issues that don’t cause an outage but that negatively impact their users’ experience.
Tolerating Non-Critical Errors is Key to a Strategic Reliability Program
SLOs operate with what’s known as an “error budget,” that is, the number of times a Service Level Indicator (SLI) fails to meet its target metric. There is no such thing as a good error, but by testing SLIs over historical data when setting up an SLO allows you to identify an acceptable error rate.
Some errors should be considered non-critical - for example, an authentication gateway that immediately tries again when an error occurs should be considered less critical than a payments API that simply stops after an error. Using SLOs with Nobl9 allows you to be strategic with your error tolerance, putting emphasis on SLIs that directly impact or impede the customer’s journey. Doing so will allow you to not only focus your efforts on the everyday user experience, but to strategically distribute your IT investments into areas that affect your real business goals.
Don’t Make Your SREs Re-Invent the Wheel
Your engineers already have their preferred tools in place to monitor and observe their particular parts of your IT infrastructure. They may have Datadog, CloudWatch, Splunk, New Relic, etc. - however they’re capturing metrics, events, logs and traces, ripping it out and replacing it is both unnecessary and likely to be met with significant pushback.
Nobl9 is platform agnostic. Your engineering teams’ existing tools can be pulled in either via one of our purpose-built integrations or by using our SLI Connect data ingestion engine. Queries can be run using the data source’s native querying language, and your Nobl9 SLO will normalize the data for an accurate, actionable single pane of glass view of what matters most to your users’ daily experience.
Making Sense of the Data
An ongoing challenge in the world of site and application reliability is actually taking meaning from the metrics. Infrastructure and application metrics are often extremely specialized, meaning that for anyone who isn’t an engineer focused on the system or service being measured may not be able to easily understand what the data actually means. Often this leads to de facto top-level metrics like nines of uptime.
Nobl9 makes it easy to understand the actual reliability of an application at a glance. Our Reliability Roll-Up Reports are incredibly useful, distilling the complexity of reliability of an application spanning a variety of systems and services into a percentage-based Reliability Score. With this, you’ll know at a glance how reliable your application actually is, without having to have a ton of technical knowledge and without oversimplifying everything into a count of nines.