Cloudian Bolsters Object Store with Monitoring and Analytics
Cloudian today rolled out a new offering called HyperIQ that it says will give customers a much deeper and wider view of their growing HyperStore object storage systems, without burdening them with another complex piece of technology.
Cloudian dubs HyperIQ an “observability and analytics” solution that’s designed to keeps tabs on the hardware underlying the HyperStore system (including the network interconnects, CPUs, and drives), as well as what the users and applications are doing on top of the object store system.
The software gives customers interactive dashboards that allow users to slice and dice current and historical data by different variables, such as resource utilization by data center or by node. Customers can also build their own custom dashboards by mixing and matching 100 available data panels.
HyperIQ wears many hats. It can be used to monitor the health and status of geographically dispersed HyperStore clusters in real-time. It can also be used for predictive maintenance and perhaps capacity planning. Security is another potential use case; if users are trying to delete protected buckets of data, for example, then HyperIQ can detect that anomalous activity.
“There’s a lot of stuff that a Cloudian administrator wants to have answered and that’s what HyperIQ helps them to do,” says Cloudian Chief Marketing Officer Jon Toor.
The new software goes above and beyond what Cloudian offers with its Management Console, which it describes as an “on box” application that is tied to the S3-based object storage system. HyperIQ, by contrast, can be spun up on a virtual machine in a matter of minutes, providing quick visibility into sprawling HyperStore cluster that can span thousands of nodes spread across multiple data centers and even countries.
“HyperIQ takes it into a completely new realm,” says Sanjay Jagad, senior director of products and solutions for the San Mateo, California based company.
“What we have here is a comprehensive infrastructure monitoring,” he continues. “When you first come in Monday morning, it gives you a complete perspective on what’s happening in your cluster. It tells you how many nodes you have, what your overall capacity that you’ve used, how much data that you’ve written, how much data that you’ve read, what are the trend lines for capacity and performance. It gives you the breakdown of capacity across different nodes. And it gives you a quick health check…for alerts.”
HyperIQ includes a built-in time-series database, which allows customers to perform historical analysis, Jagad says. That lets users compare what their system looks like and is behaving this week against the same week last year, for example. Anomalies will stand out on the screen when performance data from different time periods are shown next to each other.
As Cloudian object storage environments grow across on-premise, cloud, or hybrid-cloud environments, the systems can become more complex. The company does its best to tamp down on the complexity, but complexity inevitably creeps in when you’re storing upwards of an exabyte of data in a single global namespace with triple-way replication across thousands of nodes. Just from a hardware perspective, drives break, network cards go bad, and fans die.
Cloudian developed HyperIQ to provide a level of monitoring and visibility into these growing and increasingly complex object storage environments, but to do so without the tool itself – which uses Cassandra, Redis, Prometheus, and Grafana components under the covers – becoming overly complex and difficult to use, Jagad says.
“What this brings to the table is it gives customers an easy way to quickly get insights into their Cloudian infrastructure — especially if we’re talking about hundreds of nodes — and how can I easily manage this without having to worry about perinatal burdens,” Jagad tells Datanami.
Cloudian has hooked up the new HyperIQ offering to SmartSupport, an existing solution that brings a “phone home” aspect to Cloudian HyperStore environments. If customers consent to sending metadata collected by HyperIQ to Cloudian, the software company can analyze that data to determine how things are running on the customer system.
“You’re doing the day to day monitoring,” Toor explains. “We can join with you once a month and say, Ok, let’s look at the system over all, let’s see what kind of alerts you’re getting, see what kind of software you’re running, see that everything is up to what we consider our best practices, based on information we’re gathering from thousands of Cloudian customers worldwide.”
Cloudian is selling two version of HyperIQ, including a basic version and an enterprise version. HyperIQ Basic is included free with HyperStore licenses, and includes basic reporting and monitoring. HyperIQ brings advanced monitoring and analytics to the table at a cost of 0.025 cents per gigabyte per month, or about $3,000 per petabyte per year. SmartSupport is extra.
Cloudian, Splunk Unchain Compute, Storage
Object Stores Starting to Look Like Databases
Cloudian Sees $94M for Object Funding
Editor’s Note: This article has been corrected. Cloudian SmartSupport is not a new product. Datanami regrets the error.