DMA OperationsGuide Recommended Regular Maintenance
350 Polycom, Inc.
Every night, each Polycom DMA system cluster determines whether its
configuration or local user data have changed. If so, it creates a
configuration-only backup of the system. For details on backups, see “Backing
Up and Restoring” on page 381.
General system health and capacity checks
On the Dashboard (see “Dashboard” on page352), verify that:
•There are no alerts indicating problems with any part of the system.
•The Supercluster Status pane shows the correct number of servers and
clusters, and the network interfaces that should be working (depending on
your IP ty pe a nd sp li t ne tw or k s ett in gs ) a re u p ( gr ee n up ar ro w) and in fu ll
duplex mode, with the speed correct for your enterprise network.
•The Cluster Info pane’s Resources section shows that there is adequate
free disk space. If the system is using more than 80% of disk space, free up
space by doing some or all of the following:
—Go to Maintenance > Backup and Restore and download and delete
backup files (see “Backing Up and Restoring” on page 381).
—Go to Maintenance > System Log Files and download and delete log
file archives (you must have the Auditor role to do so; see “System Log
Files” on page 377).
—Go to Admin > Local Cluster > Logging Settings and reducing the
retention period for log archives (see “Logging Settings” on page81).
—Go to Admin > Call Server > History Retention Settings and reduce
the retention values (you must have the Auditor role to do so; see
“History Retention Settings” on page28 1).
•The Territories Status pane shows that all territories have the correct
capabilities, are being managed by their primary cluster, and (if your
deployment is so configured), have a backup cluster.
Go to Reports > Network Usage (see “Network Usage Report” on page 426)
and view the graph for each cluster with the following capacity-related metrics
selected:
• Call Counts — If the number of concurrent calls approaches the license
limit, you may need to rebalance territory responsibilities, add licensed
capacity, or add another cluster.
• Conference Manager Calls — If the number of concurrent calls
approaches the number of MCU ports available, you may need to add
MCU capacity.
View the graph for each site, site link, and subnet with Calls Dropped and
Calls Downspeeded selected. These metrics show only calls dropped or
downspeeded due to insufficient bandwidth at the selected throttlepoint. Any
values above zero are indicators of bandwidth saturation and suggest that it’s
time to increase network bandwidth.