7 viewing performance statistics, 1 overview of the collectl utility, 7 viewing performance statistics -19 – HP StorageWorks Scalable File Share User Manual
Page 75: Overview of the collectl utility -19, Viewing performance statistics (section 4.7)
Viewing performance statistics
4–19
4.7 Viewing performance statistics
Performance statistics are automatically gathered from each server in the HP SFS system. This section
describes how to view those performance statistics. The section is organized as follows:
•
Overview of the collectl utility (Section 4.7.1)
•
Overview of the information gathered by the collectl utility (Section 4.7.2)
•
Using the Web server to view information gathered by the collectl utility (Section 4.7.3)
•
Viewing information in the /proc file system (Section 4.7.4)
NOTE:
In this release of the HP SFS software, the
collectl
utility and the Web server are enabled by
default—that is, the
enable_collectl
and
enable_webserver
attributes are enabled by default.
4.7.1
Overview of the collectl utility
NOTE:
The
collectl
utility and the associated tools are provided as diagnostic tools, for use mainly by
HP Service personnel.
The
collectl
utility and the associated tools are provided without warranty for HP SFS customers.
The
collectl
utility is available for certain client distributions. Contact your HP Customer Support
representative to find out if it is available for your particular system.
The
collectl
utility gathers performance metrics from each server in the HP SFS system. By default, the
collectl
utility is enabled on the HP SFS system; that is, the
enable_collectl
attribute is set to
yes
.
If the utility has been disabled, you can enable it again by entering the following commands:
sfs> set attribute enable_collectl=yes
sfs> configure server all
When the
configure server all
command has finished executing, all of the servers will be running
the
collectl
daemon.
The
collectl
utility runs as a low-overhead system daemon; when it is enabled on the HP SFS system, it
is started automatically at boot time on each server. Every five minutes, the performance metrics are collated
on the administration server by the
/usr/opt/hpls/diags/bin/collate_perf_stats.bash
script; when the Web server is enabled, the script also processes all of the performance logs so that they
can be viewed on the Web server
The logs that the
collectl
utility generates are automatically gathered on the administration server every
five minutes; in addition, you can manually update the logs by entering the following command on the
administration server:
# /usr/opt/hpls/diags/bin/collate_perf_stats.bash
By default, performance data is maintained for the previous 20 days. You can change the number of days
by changing the value of the
plot_history
attribute, as shown in the following example:
sfs> set attribute plot_history=15
This change takes effect immediately; no other action is required.
NOTE:
On systems where the administration LUN is smaller than 10GB, the
plot_history
setting is
ignored, and only four days of history are kept.