Wouldn’t it be great if you were able to have a better understanding of your data storage performance so you could quickly and adjust workloads accordingly? Kaminario customers are taking advantage of this capability with Kaminario Clarity. Clarity is a SaaS-based predictive analytics platform that delivers intelligence, automation, and analytics to customers’ cloud-scale infrastructure. In this installment of a video series highlighting the capabilities of Clarity, we’ll demonstrate how Clarity helps customers analyze performance loads and plan how to reallocate resources based on latency and bandwidth needs.
With Kaminario Clarity, users can quickly understand where performance loads reside and how they can be reallocated. Data can easily be filtered at both the volume and volume group level so users can quickly trace down any possible performance issues, determine what resources are creating the highest latency, and rebalance based on bandwidth constraints.
Predictive Analytics for Post-Advanced Analysis
Kaminario Clarity allows users to look at the latency, IOPS, and bandwidth of volumes or volume groups in a single array during a particular timeframe (whether that’s days, weeks, months, or a custom time period). But even more interesting than looking at the historical performance data of the array is Clarity’s advanced analysis data.
For the user who is curious about what would happen to performance in a particular scenario (i.e. an ecommerce website during the holiday season), Kaminario Clarity enables predictive modeling. Based on historical data and/or specific thresholds, Clarity offers users more granular control over predictive performance analytics than ever before.
Deep-Diving into the Performance of Volumes and Volume Groups
You can dive even further into this data for individual volumes or volume groups. IOPS, throughput, and latency from volume groups and volumes can be sorted and selected in a table. When looking at individual volumes, users can also identify if data has been deduped and what the allocated capacity is. In the demo video above, we determine that VD-read04 tends to be busier than other volumes. To reduce its maximum latency, we could move it to a different volume or split up the volume groups. If it was deduced and contained large writes, we could move it to a non-deduced volume or volume group. In addition, users also can sort data based on block size or to only look at reads or writes.
Sharing Data from Kaminario Clarity
The table and chart data from Kaminario Clarity can be exported to Excel and used in presentations or pie charts. Alternatively, you can export to Excel to cut data and review it in a granular state other than the GUI.