Monitoring I/O performance
Flex Appliance collects disk I/O metrics for the LUNs that back your appliance storage and for the volumes used by your application instances. Disk I/O metrics are available on the page's tab.
Metrics are also available through the existing Flex metrics API, and you can pull them into an external monitoring stack.
You can see how the storage on those LUNs is allocated to your instances rather than how it is being used.
See Viewing instance storage allocation across LUNs.
The tab has a single dropdown with two modes on the same page:
Appliance LUNs (default) - It shows platform-level LUN I/O.
See Appliance LUN I/O.
Instance volumes on LUN - It shows I/O for the instance volumes on a specific LUN.
I/O metrics are kept for up to 30 days.
Charts do not refresh automatically. Data reloads when you open the tab, click , change the time frame, or switch the dropdown.
I/O counters reset to zero when a node reboots. You will see a gap or a drop to zero in the metrics immediately after a restart.