(the%b and svc_t columns, respectively). Service times are unimportant for disks that are less
thanabout 30% busy, but for busier disks, service times should not exceed about 20
milliseconds.If your busy disks have slower service times, improving disk performance might
helpWeb Server performance substantially.
Yourrst step should be to balance the load: if some disks are busy while others are lightly
loaded,move some les o of the busy disks and onto the idle disks. If there is an imbalance,
correctingit usually gives a far greater payo than trying to tune the overloaded disks.
Solaris Platform-Specic PerformanceMonitoring
Thissection describes some of the Solaris-specic tools and utilities you can use to monitor
yoursystem's behavior, and includes the following topics:
■“Short-TermSystem Monitoring” on page 94
■“Long-TermSystem Monitoring” on page 95
■““Intelligent”Monitoring” on page 95
Thetools described in this section monitor performance from the standpoint of how the system
respondsto the load that Web Server generates. For information about using Web Server's own
capabilitiesto track the demands that users place on the Web Server itself, see “Monitoring
ServerPerformance” on page 22.

Short-TermSystem Monitoring

Solarisoers several tools for taking “snapshots” of system behavior. Although you can capture
theiroutput in les for later analysis, the tools listed below are primarily intended for
monitoringsystem behavior in real time:
■Theiostat -x 60 command reports disk performance statistics at 60-second intervals.
Watchthe %b column to see how much of the time each disk is busy. For any disk busy more
thanabout 20% of the time, pay attention to the service time as reported in the svct column.
Othercolumns report the I/O operation rates, the amount of data transferred, and so on.
■Thevmstat 60 command summarizes virtual memory activity and some CPU statistics at
60-secondintervals.
Monitorthe sr column to keep track of the page scan rate and take action if it's too high
(notethat "too high"is very dierent for Solaris 8 and 9 than for earlier releases). Watch the
us,sy,and id columns to see how heavily the CPUs are being used; remember that you need
tokeep plenty of CPU power in reserve to handle sudden bursts of activity. Also keep track
ofthe rcolumn to see how many threads are contending for CPU time; if this remains
higherthan about four times the number of CPUs, you might need to reduce the server's
concurrency.
SolarisPlatform-SpecicPerformance Monitoring
SunJavaSystem Web Server 7.0 Update 1 Performance Tuning,Sizing, and Scaling Guide •94