Pivot3 Acuity includes predictive analytics specific to the following areas:
- Proactive supportability: sensor data from multiple components (flash, disk, node, etc) within the system is analysed for predictive failures and self-healing of the system. For instance, if a drive is failing (not yet failed), system will automatically spare it out and phone home order a new one. Also, we analyse and predict when flash wear-out may occur and display/alarm this information to the customer and phone home.
- Capacity and Performance Planning: complete suite of metrics and dashboards available to the customer for capacity (current and predictive future) and performance to aid with expansion planning. All these metrics are included in a daily phone-home package sent to Pivot3 support cloud.
- Realtime IO Path/Data Placement manipulation: every IO is tracked and analysed in order to meet the SLAs defined by the customer via the built-in policy engine. Data placement in the system (RAM, NVMe, SSD, HDD) is governed by policy. The system uses predictive analytics and automation to make sure right IO path queueing and data placement is in effect to meet SLAs. Policy changes can be automated via built-in scheduler and CLI/APIs.