DEVOPSdigest

To reduce the time it takes for engineering teams to detect an issue with a feature release, Split Software announced the addition of Feature Monitoring, an automated detection capability for its feature delivery platform that reduces detection times of errors in a code release. This allows teams using feature flags to react to issues much faster and use an “instant off” option within five minutes, preserving user experience without the need for rollbacks or hot fixes.

“Engineering teams are spending too much time detecting issues with new releases. In fact, we took a survey at three recent DevOps conferences and found that among the professionals we spoke to, a majority – about 70 percent – said it can take more than an hour for their organizations to detect issues with new releases, while about a third said it takes more than 24 hours for their teams to do the same. Although just a snapshot, it demonstrates the larger issue of the costly amount of time wasted, especially as teams strive to deploy faster and with less risk,” said Adil Aijaz, co-founder and CEO, Split. “However, using Split, customers are able to more quickly detect errors or performance degradation – within a matter of minutes -- as well as immediately remediate the error, before a customer even notices a problem. By removing the risk of service disruptions, Split is helping to unlock innovation.”

Split is removing considerable risk from the deploy process by connecting errors to specific features. Currently, no other feature flagging system matches features to the errors they are causing, which means it can take time to determine the root cause and this opens the app and the company to the risk of customer complaints. However, Feature Monitoring automatically detects a misbehaving feature. This means that teams can use a kill switch based on a feature flag to shut down a non-working or unnecessary feature before it negatively impacts a customer’s experience.

Feature Monitoring empowers businesses of any size to release features quickly and confidently, and is one of several best practices that Split builds into its feature delivery platform, including:

- A secure way to release new features

- The ability to target features to customers with precision

- Measurement of the impact of features on customer experience metrics

- Feature flags to separate code deploy from feature release and safely execute continuous delivery at scale

- Observability of key performance indicators (KPIs) like functionality and performance tied to features.

Split’s Feature Monitoring also includes pre-built integration with Sentry, to capture error data for each feature, and includes a Lightweight JS agent alongside Split’s JavaScript SDK to capture real user performance data.