Section 9.6. Interrupt Statistics: vmstat -i

9.6. Interrupt Statistics: vmstat -i

Another useful measure of kernel activity is the number of received interrupts. A device may be busy processing a flood of interrupts and consuming significant CPU time. This CPU time may not appear in the usual by-process view from prstat.

The -i option of the vmstat command obtains interrupt statistics.

$ vmstat -i interrupt         total     rate -------------------------------- clock         272636119      100 hmec0            726271        0 audiocs               0        0 fdc0                  8        0 ecppc0                0        0 -------------------------------- Total         273362398      100 

In this example, the hmec0 device received 726, 271 interrupts. The rate is also printed, which for the clock interrupt is 100 hertz. This output may be handy, although the counters that vmstat currently uses are of ulong_t, which may wrap and thus print incorrect values if a server is online for several months.