Help

Table Of Contents
Check if CPU utilization and front-end operations count is also high. The system might be simply at over-capacity, which
may also be reflected in this category.
Constantly rising heap usage is a potential sign of a director memory leak. Contact Dell EMC Support to identify these
systems.
Corrective actions
Check if CPU utilization and front-end operations count is also high. The system might be simply at over-capacity, which
may also be reflected in this category.
Constantly rising heap usage is a potential sign of a director memory leak. Contact Dell EMC Support to identify these
systems.
Changing the view
To view the Heap usage of a single director in your metro node system, select the director name from the Director drop-down.
NOTE: The chart displays data only for the directors in the cluster to which you are currently connected. To simultaneously
view statistics for another cluster, open a second browser session and connect to the second cluster.
Viewing the Heap Usage chart
1. From the GUI main menu, click Performance.
2. In the Performance Dashboard, select the tab in which you want to display the chart (or create a custom tab).
3. Click +Add Content.
4. Click the Heap Usage chart icon.
Front-end Queue Depth chart
The Front-end Queue Depth chart provides the count of front-end operations per director. It describes the number of
concurrent outstanding operations active in the system. This statistic is not directly related to front-end IOPS. IOPS could
be easily be in the thousands range, whereas this field describes the concurrency of the operations and may be quite low,
depending upon how many I/O's the application or host have outstanding to metro node over a given time.
Guidelines
Keep in mind the following guidelines when using the Front-end Queue Depth chart:
Normal values are approximately 10-100 concurrent operations for a small to medium system.
Anything greater than about 500 shows a lot of concurrent access in a system but is not alarming. Check the CPU
utilization.
> 1000 active could be indicative of a problem with metro node keeping up with host I/O demands. Check CPU utilization,
front-end aborts, and be-aborts
This value depends heavily upon the customer workload.
Corrective actions
Monitor CPU % busy, front-end aborts, back-end aborts, and WAN health state (in metro node Metro systems).
If other directors in a cluster are less busy, look to balance the workload across these directors.
SCSI CDB trace information can identify if there are large dormant queues in the front-end. This can be bad for
performance.
For metro node Metro systems, WAN performance degradation could cause active operations to distributed-devices to not
complete as quickly as they did previously. This may result in an I/O operations queue build up, causing this counter to
increase. Check the WAN connectivity, health state, and performance.
In VMware and certain host clustered environments, SCSI reservations could causing spikes in active operations, if there are
a lot of reservation-intensive operations (vMotion, Storage vMotion, backup, or frequent file creations/deletions).
38
Monitoring the system