For many companies, cloud costs are among the top investments these days. With a growing number of services, instances and regions, cloud cost optimization is becoming increasingly painful. Companies use cloud management platforms to optimize costs and increase cloud visibility and security. But staying on top of AWS budgets requires proficiency, agility and time—especially when any glitch can result in massive cost bleeds.

However, cloud cost monitoring and cloud cost optimization are not the same thing. That’s why leading companies are adding another layer on top of cloud cost management and optimization, using AI-based ‘Real-time Usage Alerts’ for proactive monitoring of their cloud management stack. While optimization will help reduce costs in the long term, real-time cloud cost monitoring ensures lightning-fast detection of the incidents that impact your revenue so that your spend is always on track with no surprises in your monthly bill.

Companies experience a slew of cloud incidents over the course of a month. A break, even on a granular level, can easily rack up thousands of dollars. In medium to large companies, cloud glitches, undetected for a course of a day, set back the company $5K-$50K on average. A good example can be found in a cloud cost anomaly experienced by Copilot.cx, a leading automated Customer Experience platform, detected in real-time by Anodot. The anomaly was found in Google BigQuery, when a bug in the system caused many more queries than normal to run, causing the cost to rise by more than $199 per hour, which would have resulted in a minimum $4,800 loss — If not for Anodot’s early detection cost alerts.

AI models can provide proactive cost control by catching unusual spikes in real time. Now, Anodot is set to eliminate surprises in your cloud invoices with machine learning-based detection & real-time alerts for AWS. Anodot’s anomaly detection capabilities catch runaway AWS cloud spend immediately, alert appropriate teams, and provide deep root-cause analysis so that problems can be remediated quickly. 

 

Cloud Cost Monitoring 1

AWS cost monitoring made easy

Even though AWS and cloud management platforms provide daily cost reports, in most cases this isn’t enough. On the one hand, the daily report is very busy, with lots of data that can’t be monitored. On the other hand, it doesn’t provide the granularity that’s required by real-time monitoring. For example, incorrectly querying a database for a few hours can cause costs to skyrocket. With a daily report, you wouldn’t be able to detect the spike until it’s too late. Analyzing the report, in depth, with anomaly detection and real-time alerts is the smart way to monitor daily costs and avoid acute bleeds that negatively impact the bottom line.

Anodot makes it easy to stay on top of AWS costs by using autonomous monitoring that is based on proprietary anomaly detection capabilities. Anodot collects all AWS metrics on an hourly basis and CUR files to develop comprehensive costs and usage dashboards and demand forecasts. The comprehensive AWS cost monitoring pack helps reduce costs immediately with its three-layer business package:

Real time cost & usage monitoring

Anodot monitors AWS services usage and costs on an hourly basis with a high level of granularity: costs and usage are specific to the service, region, team, and instance type. This means that if usage spikes, you don’t need to wait a full day to resolve the issue and can actively stop cost increases in their tracks. When anomalies do occur, this level of granularity allows for a much faster time-to-resolution.

Comprehensive cost reports

Anodot’s anomaly detection uses autonomous analytics to monitor your AWS CUR files in real time and provides you with highly granular and comprehensive cost and usage dashboards.

Forecast future costs

Anodot learns the normal behavior of every service and runs prediction models based on historical CUR files to generate accurate forecasts for future AWS demand. These allow for more effective budget planning and resource allocation.

Good Catch! Real Customer Examples

Spike in EC2 CPU Count

Spike in EC2 CPU Count

The above alert is from an AWS EC2 service monitored on an hourly time frame. In this case, the alert was set for a spike in CPU count over 3 hours in duration. Total CPU counts can go up and down frequently according to application demands, but if you’re looking at the current cycle you often don’t actually know how much you’re using relative to normal levels.

The key difference in this example is that, instead of simply looking at absolute values, Anodot’s anomaly detection was looking at the average hourly CPU count. It’s clear that the spike in CPU count is not larger than the typical spikes: what is anomalous is the time of day of the spike. In this case, by looking at the average hourly CPU count and monitoring on a shorter time frame, the company received a real-time alert and was able to resolve the problem before it incurred a significant cost.

Spike in EC2 Network Traffic

Spike in EC2 Network Traffic

In this example, we can see that the service is an AWS EC2 instance, which is being monitored for network traffic on an hourly basis. The service experienced a +339% increase in incoming network traffic over the course of three hours. With the real-time alert sent to the appropriate team, which was served with a root-cause analysis, the anomaly was promptly resolved, ultimately resulting in cost-savings for the company.

Spike in CloudFront Requests

Spike in CloudFront Requests

A final example is this alert for AWS CloudFront service, which was again being monitored on an hourly timescale. In this case, there was an irregular spike in the rate of CloudFront requests for an S3 bucket. Similar to other examples, if the company was only monitoring costs reactively at the end of the day, this could have severely impacted the bottom line. By taking a proactive approach to cloud-cost monitoring with the use of AI and machine learning, the anomaly was quickly resolved and the company was able to avoid otherwise wasted costs.

Stop cloud cost spikes in their tracks

Anodot catches runaway cloud spend immediately, alerts appropriate teams, and provides deep root-cause analysis across all your cloud resources. Rein-in your cloud costs with real-time autonomous monitoring and optimization. Forward-thinking teams use Anodot to avoid surprises in their AWS bill. You too can get started with a free trial of Anodot’s real-time AWS cost alert & forecasting solution right here or at AWS marketplace.

 

 

 

Topics: AnalyticsBusiness IntelligenceAnalytics|Machine Learning|Big Data|Anomaly DetectionCloud Cost MonitoringAmazon Web ServicesAWSCloud Costs
Subscribe to our newsletter and get the latest news on AI analytics

Written by Yuval Dror

Yuval Dror is Director of Engineering and Head of DevOps at Anodot.

This website stores cookies on your device. These are used to improve your website experience and provide more personalized services. Learn more