This post explains how to obtain virtual or physical CPU usage of an Oracle VM Server, including actual CPU usage and running time of guest domains. Guest-specific (Dom-U) CPU usage can be obtained by running commands on the guest itself, for example; top, ps, and sar.
If, however, you wish to obtain specific CPU usage, including that of the Oracle VM Server management domain/hypervisor (Dom-0), the following xen-tools utilities are available to run on the Oracle VM Server:
- /usr/sbin/xentop -v
- /usr/sbin/xenmon.py
1. xentop
The xentop -v command display virtual CPU information e.g.:
# xentop -v
6 domains: 2 running, 3 blocked, 0 paused, 0 crashed, 0 dying, 0 shutdown
Mem: 8312440k total, 6212448k used, 2099992k free CPUs: 2 @ 2128MHz
NAME STATE CPU(sec) CPU(%) MEM(k) MEM(%) MAXMEM(k) MAXMEM(%) VCPUS NETS NETTX(k) NETRX(k) VBDS VBD_OO VBD_RD VBD_WR SSID
4461_EL5U3 ------ 53060 0.0 1056636 12.7 1064960 12.8 1 0 0 0 3 0 0 0 0
VCPUs(sec): 0: 53060s
4474_EL5U2 ------ 45070 0.0 1056636 12.7 1064960 12.8 2 0 0 0 2 0 0 0 0
VCPUs(sec): 0: 32213s 1: 12857s
4516_EL4U6 -----r 512039 0.0 1056636 12.7 1064960 12.8 1 0 0 0 2 0 0 0 0
VCPUs(sec): 0: 512039s
4570_EL464 --b--- 302 0.0 2097152 25.2 2097152 25.2 1 1 10 218297 1 0 22371 16768 0
VCPUs(sec): 0: 302s
Domain-0 -----r 87787 0.0 254976 3.1 no limit n/a 2 8 1717873 4158394 0 0 0 0 0
VCPUs(sec): 0: 57819s 1: 29968s
4511_EL564 --b--- 108 0.0 524288 6.3 524288 6.3 2 1 342 214623 1 0 822 1 0
VCPUs(sec): 0: 96s 1: 12s
In the example above, domains Domain-0, 4474_EL5U2 and 4511_EL564 each utilize two virtual CPUs, whereas all others only utilize one virtual CPU. Domain Domain-0 utilizes a total CPU time of 87787 seconds; 57,819 seconds on VCPU0 and 29,968 seconds on VCPU1.
Xentop usage:
Usage: xentop [OPTION] Displays ongoing information about xen vm resources -h, --help display this help and exit -V, --version output version information and exit -d, --delay=SECONDS seconds between updates (default 3) -n, --networks output vif network data -x, --vbds output vbd block device data -r, --repeat-header repeat table header before each domain -v, --vcpus output vcpu data -b, --batch output in batch mode, no user input accepted -i, --iterations number of iterations before exiting -f, --full-name output the full domain name (not truncated) # Note: this can be used from the latest version of Oracle VM Server 2.2.2
The parameters mean:
| Parameters | meanings |
|---|---|
| CPU(sec) | CPU time which the guest OS has consumed(cumulated) |
| CPU(%) | “CPU percentage which the guest OS consumes currently. Note1: This is calculated as CPU(%) = Actuall Assigned CPU# * Actual CPU Usage on the Guest OS(0-100%).Thus this could be up to 300% when 3 CPUs are assigned to the Guest OS. “”Actuall Assigned CPU#”” is not the Virtual CPU Number assigned to the Guest OS, But Actual CPU Number assigned to the Guest OS, since it should be up to 200% when physical CPU cores are 2 even if VCPUs assigned to the Guest OS are 3 or more. Note2: this is always 0.0% at the 1st xentop output, since this is calculated by dividing the interval of the updates. Thus “”-i”” option needs to be set as 2 or more to get correct CPU(%). Note3: This just shows the CPU usage on Host OS for the Guest OSs, therefore this could be different from the CPU usage (ex. which “”top”” shows) on the Guest OSs.” |
| MEM(k) | Memory size in Kbyte which the guest OS occupies on host OS memory. For Domain0, this is equal value of dom0_mem= specified in /etc/grub.conf of Host OS, and also equal value of “memory =” in vm.cfg for Guest OS. ex. dom0_mem=1024M shows 1048576(kB) for MEM(k) for the Host OS. |
| MEM(%) | Memory size in percentage which the guest OS occupies on host OS memory |
| MAXMEM(k) | Max memory size in Kbyte which the guest OS could occupies on host OS memory |
| MAXMEM(%) | Max memory size in percentage which the guest OS occupies on host OS memory |
| VCPUS | Virtual CPU number which is assigned to the guest OS |
| NETS | Virtual network interface number which is assigned to the guest OS |
| NETTX(k) | network traffic in Kbyte the guest OS sends(cumulated) |
| NETRX(k) | network traffic in Kbyte the guest OS receives(cumulated) |
| VBDS | Virtual block device(=storage) number which is assigned to the guest OS |
| VBD_OO | blocked number of I/O from/to the guest OS |
| VBD_RD | read request number from the guest OS(cumulated) |
| VBD_WR | write request number from the guest OS(cumulated) |
| VBD_RSECT | read sector number from the guest OS(cumulated), old xentop does not show this |
| VBD_WSECT | write sector number from the guest OS(cumulated), old xentop does not show this |
| SSID | N/A |
2. Xenmon.py
The xenmon.py command displays specific physical CPU information e.g.:
# /usr/sbin/xenmon.py CPU = 0 Last 10 seconds (99.80%) Last 1 second (99.72%) =============================================================================================================================================================================== 1 13.98 ms 1.40% 14.57 us/ex 10.40 ms 1.04% 14.00 us/ex Gotten 1 456.93 ms 45.69% 0.00 ns/io 349.26 ms 34.93% 0.00 ns/io Blocked 1 486.25 ms 48.63% 506.85 us/ex 36.53 ms 3.65% 49.15 us/ex Waited 0 40.33 ms 4.03% 101.06 us/ex 69.70 ms 6.97% 177.92 us/ex Gotten 0 588.45 ms 58.84% 0.00 ns/io 581.26 ms 58.13% 0.00 ns/io Blocked 0 17.87 ms 1.79% 44.77 us/ex 37.90 ms 3.79% 96.75 us/ex Waited 28 26.00 ms 2.60% 16.02 us/ex 20.30 ms 2.03% 14.22 us/ex Gotten 28 611.88 ms 61.19% 0.00 ns/io 586.71 ms 58.67% 0.00 ns/io Blocked 28 26.19 ms 2.62% 16.14 us/ex 67.13 ms 6.71% 47.02 us/ex Waited Idle 448.98 ms 44.90% 229.59 us/ex 359.74 ms 35.97% 251.82 us/ex Gotten Idle 0.00 ns 0.00% 0.00 ns/io 0.00 ns 0.00% 0.00 ns/io Blocked Idle 550.66 ms 55.07% 281.58 us/ex 640.01 ms 64.00% 448.01 us/ex Waited 41 1.40 ms 0.14% 54.30 us/ex 964.77 us 0.10% 57.33 us/ex Gotten 41 459.20 ms 45.92% 0.00 ns/io 470.01 ms 47.00% 0.00 ns/io Blocked 41 1.94 ms 0.19% 75.09 us/ex 11.80 ms 1.18% 701.16 us/ex Waited 13 466.75 ms 46.67% 863.83 us/ex 536.81 ms 53.68% 1.06 ms/ex Gotten 13 0.00 ns 0.00% 0.00 ns/io 0.00 ns 0.00% 0.00 ns/io Blocked 13 8.52 ms 0.85% 15.77 us/ex 8.46 ms 0.85% 16.75 us/ex Waited 49 568.29 us 0.06% 34.60 us/ex 444.77 us 0.04% 31.72 us/ex Gotten 49 546.79 ms 54.68% 0.00 ns/io 560.84 ms 56.08% 0.00 ns/io Blocked 49 1.13 ms 0.11% 69.04 us/ex 6.18 ms 0.62% 441.01 us/ex Waited * 99.80% 99.84%
The above example output displays information about physical CPU 0 i.e. how CPU=0 is used by various domains and how long it has been used within the last 1 and 10 seconds. The first column reports the Domain ID – 0 refers to Dom-0.
Three lines are printed for each domain:
- 1st row: Gotten = amount of the CPU time/percentage which the CPU is running on the domain.
- 2nd row: Blocked = amount of sleeping time/percentage.
- 3rd row: Waited = amount of the CPU time/percentage which the domain is in wait state.
Press the ‘p’ or ‘n’ key to display information for the previous/next physical CPU.
Xenmon.py usage:
usage: xenmon.py [options]
options:
-h, --help show this help message and exit
-l, --live show the ncurses live monitoring frontend (default)
-n, --notlive write to file instead of live monitoring
-p PREFIX, --prefix=PREFIX
prefix to use for output files
-t DURATION, --time=DURATION
stop logging to file after this much time has elapsed
(in seconds). set to 0 to keep logging indefinitely
-i INTERVAL, --interval=INTERVAL
interval for logging (in ms)
--ms_per_sample=MSPERSAMPLE
determines how many ms worth of data goes in a sample
--cpu=CPU specifies which cpu to display data for
--allocated Display allocated time for each domain
--noallocated Don't display allocated time for each domain
--blocked Display blocked time for each domain
--noblocked Don't display blocked time for each domain
--waited Display waiting time for each domain
--nowaited Don't display waiting time for each domain
--excount Display execution count for each domain
--noexcount Don't display execution count for each domain
--iocount Display I/O count for each domain
--noiocount Don't display I/O count for each domain