Help

Changing the view
Use the following appropriate selection criteria to filter the data:
● Director — Allows you to select all directors or a specific director in the cluster.
● Read and Write check boxes — Allows you to select one or both check boxes to filter throughput for Reads and Writes.
Viewing the Back-end Throughput chart
1. From the GUI main menu, click Performance.
2. In the Performance Dashboard, select the tab in which you want to display the Back-end Throughput chart (or create a
custom tab).
3. Click +Add Content.
4. Click the Back-end Throughput icon.
Back-end Errors chart
The Back-end Errors chart displays the back-end I/O errors to and from the storage array. There are three categories of
back-end errors:
● Aborts — Indicate the metro node back-end gave up and aborted the I/O operation to the storage array, the array itself
decided to abort the I/O operation, or another SCSI initiator (metro node director, or host) connected to the array caused
the I/O to abort.
● Timeouts — Indicate the metro node back-end saw an I/O operation to a storage volume that has not completed within 10
seconds.
● Resets — Logical Unit resets issued by the metro node back-end to a storage volume as corrective action when after 20
seconds of no response for any I/O by the storage volume (Logical Unit on the storage array). The metro node back-end
re-tries all outstanding I/O to the storage volume.
NOTE:
The chart displays data only for the cluster to which you are currently connected. To simultaneously view back-end errors
for another cluster, open a second browser session and connect to the second cluster.
Guidelines
● Back-end errors typically indicate back-end fabric and/or storage array issues.
● For a normal healthy system, there should be no aborts, timeouts, or resets.
○ Timeouts might happen during bursts of I/O to a storage array. Seeing a few of these is generally not bad for
performance, however, frequent or periodic timeouts are not normal.
○ Aborts and Resets likely indicate major performance issues on the storage fabric or storage array.
○ Investigate the cause for back-end errors immediately.
Corrective actions
● Look closely at the latency related categories (front-end read/write latency, and back-end read/write latency) for any high
averages or large spikes. Try to correlate the spikes to the errors.
● Examine the back-end fabric for changes, reported errors, proper negotiated speeds, and health state.
● Examine the back-end storage array for general health state, and best practices disk/volume layout.
● Check the metro node firmware log for events indicating command timeouts, retries, or other general back-end health
issues.
Changing the view
Use the following appropriate selection criteria to filter the data:
● Director — Allows you to select all directors or a specific director in the cluster.
34
Monitoring the system