The System Overview panel is the highest-level
monitoring panel in SAN File System. From this panel you can monitor major
system components, including cluster state, clients, filesets, servers, hardware,
and recent events.
Attributes
The Cluster section
of the System Overview provides the following details:
Cluster state |
Master Workload Transactions |
Active Clients The number of clients that currently have one or more sessions. Tip: Active Clients does not represent the number of total client sessions,
which is nearly always the number of active clients multiplied by the number
of servers.
|
Actions
The Panel
control section of the System overview allows you to control the panel
refresh interval.
Refresh Interval Disables the refresh interval or sets the interval to a value of one,
two, five or fifteen minutes. The refresh occurs when the items monitored
by the System Overview panel are active. |
Submit Submits the refresh interval that was specified. Note: This selection
is reflected upon the next panel update, not necessarily when you click Submit,
unless the panel update interval exceeds the chosen refresh interval.
|
Fields
The Activity
and Availability per Server section of the System Overview panel lists
the following details:
Aggregate status icon |
Server The list of all servers in the cluster. Each server name links to additional metadata server details. The master metadata server is indicated by the
 icon. |
Server State |
Filesets The number of filesets being handled by the associated metadata server. The number in this field is also a link to additional fileset
details. The accompanying bar graphs represent how the fileset load is
distributed among the metadata servers.
For example, the metadata server with the most filesets, has the longest bar.
The bar graphs decrease in length as the number of filesets per metadata server decreases. |
Fileset Alerts The alert level for the filesets on the associated metadata server.
The alert can be one of the following levels: -
Error
— Indicates that there are one or more hard-quota violations.
-
Warning
— Indicates that there are not any hard-quota violations, but one or more
usage (Used %), or soft quota violations.
-
Normal
— Indicates that there are no fileset alerts. All filesets are below quotas
and usage (Used %) alerts.
Also listed is the total number of filesets that surpass one of the quota-related
thresholds (percentage used, soft quota, or hard quota). The
number in this field is also a link to additional fileset details. For each fileset, the statistics are given for any error or warning-level
alerts. |
Trans (per min) The number of fileset transactions, per minute, for the server. |
Load Share (%) The associated metadata server's
percentage share of the total load on all the metadata servers. The accompanying bar graphs represent how the load is distributed
among the metadata servers. For
example, the metadata server with
the largest load share percentage has the longest bar. The bar graphs decrease
in length as the load share percentage per metadata server decreases. |
Fields
The Recent
Messages section of the System Overview panel displays the ten most recent
messages that are in the cluster log, and that match the Message Types setting
for this section.
Actions
Levels Filters can be set for the following alert levels that appear in the Recent
Messages table: - Severe
- Error and above
- Warning and above (default)
- All levels, including Information
|
Types Sets a filter of All Messages or Events Only. Tip: An event is
a significant or major message that will often, but not always, trigger an
alert.
|
Submit Submits the choice for message type. Note: This selection is reflected
upon the next panel update, not when you click Submit,
unless the automatic refresh interval is now less than the elapsed time since
the last update.
|
Fields
Message ID Each message follows this convention: “XXX” Component | “YY” Subcomponent | “nnnn” Message code | “Z” Severity (Severe, Error, Warning, or Information level) |
|
Level |
Type The messages type, which is one of the following categories: - Normal — Messages that are not events.
- Event — A significant or major message that can often, but not always,
trigger an alert.
- Audit
- Trace
|
Server The metadata server name that
is the source of the message. |
Date and Time The system (not local) time that the message was issued. |
Message The text of the message. |