As the cloud world increases in complexity day by day, the majority have realized that it’s not just a necessity to monitor your cloud portfolio, but have automated mechanisms which can help you govern them better and avoid costly surprises. Anomaly detection now has made its place in the cloud world.
The general definition of anomaly is an abnormality, an irregularity, a deviation from the common rule – anomaly detection has become an effective means to help cloud platform administrators improve their cloud reliability while they constantly monitor and analyze cloud behavior. This also helps them take proactive actions and avoid system failures or crash.
What are Negative Anomalies?
While it’s observed that most reported cost anomalies are sudden increases in costs, administrators are reporting that a sudden decrease is equally damaging, if not more. Some of the common examples causing a sudden decrease or a negative anomaly are:
- A machine or service stopped working suddenly
- Cost price changed for the service
- Outage or major issue with CSP
- Misconfiguration or human error resulting in downward trend
However, on the flip side, one may also use this as a validation of certain cost optimization measures taken on the cloud platform.
The challenge though remains that while many platforms are present which help detect sudden increase in cost, there are limited which can help you detect sudden decrease in cost of cloud services.
How VMware Aria Cost powered by CloudHealth Can Help
Since the launch of our anomaly detection feature, we have received many requests about implementing negative anomalies. At VMware Aria Cost, we highly value customer feedback and excitedly introduced the detection of negative anomalies to start identifying any sudden decreases in cost.
It was observed that out of all anomalies being identified for any given account, around 15% were indeed negative anomalies i.e., anomalies which had sudden decrease in cost (and some even had far greater cost impact than multiple increase in cost anomalies). Most identified cost anomalies are associated with the usage, and we did identify few reasons behind these sudden drops in cost
- Maybe someone disabled backups
- Maybe someone bought reservations
- Maybe someone changed VM sizes
Our negative anomaly feature is helping cloud owners take charge of what is changing in their cloud portfolio and define accountability and ownership better than before.
To ensure easy usage of negative anomaly detection, users of VMware Aria Cost have multiple filters to identify anomalies such as account/subscription, region, service etc., to focus on specific areas and even filter cost impact to focus on anomalies based on impact. You not only get the ability to see detected anomalies, but also have the ability to identify the root cause of the occurring or occurred anomaly. A user can save certain filters and create a report and subscribe to the same to get notified in case of an anomaly in the portfolio.
As we receive more feedback and identify more customer use cases, we at VMware Aria Cost powered by CloudHealth, will continue our journey to come up with more pathbreaking features to assist cloud owners own their cloud.
