YotaScale Anomaly Detection
Turn weeks into minutes! Stay in control and gain time to solve more problems with tools that give dev-ops engineering teams the edge in productivity . Algorithms identify opportunities to improve performance, utilization and reduce cost. They explain why problems are happening so you can fix them fast. Cut down on errors and rework. Improve workflow efficiency. And, because it’s AI powered, the algorithms get better all the time as they learn your environment ensuring contextually customized dev-ops support.
Here’s a before and after example.
Before: Your AWS footprint is significant. You notice a big cost jump in one of your AWS accounts and realize the issue has been going on for over two weeks. Your VPE and the DevOps lead spend a week trying to figure out what was caused the spike. You load the billing files (more than 1 GB / day) into a DB for analysis. You write queries to pull a specific dataset into a spreadsheet for further processing. Now that you’ve isolated the issue to specific instance types and days, you pull additional inventory data to figure out which owner, application and resource created this anomaly. Now you connect system CPU, disk and memory metrics to analyze the root cause. Time elapsed: 3 weeks.
After: YotaScale pinpoints the issue immediately – an ELB which provisioned a number of c3.4xlarge instances because of a policy change for your analytics application. YotaScale identifies the IAM/resource owners. Time elapsed: 3 minutes.