esxtop help
Esxtop version 3.1.0
Secure mode Off
Esxtop: top for ESX
These single-character commands are available:
^L - redraw screen
space - update display
h or ? - help; show this text
q - quit
Interactive commands are:
fF Add or remove fields
oO Change the order of displayed fields
s Set the delay in seconds between updates
# Set the number of instances to display
W Write configuration file ~/.esxtop3rc
e Expand/Rollup Cpu Statistics
V View only VM instances
Sort by:
U:%USED R:%RDY N:Default
Switch display:
m:ESX memory d:ESX disk adapter u:ESX disk device v:ESX disk VM n:ESX nic
c:ESX cpu
ESX cpu
Current Field order: ABCDEfgh
* A: ID = Id
* B: GID = Group Id
* C: NAME = Name
* D: NWLD = Num Members
* E: %STATE TIMES = CPU State Times
F: EVENT COUNTS/s = CPU Event Counts
G: CPU ALLOC = CPU Allocations
H: SUMMARY STATS = CPU Summary Stats
ESX memory
Current Field order: aBCDefGHijklMno
A: ID = Id
* B: GID = Group Id
* C: NAME = Name
* D: NWLD = Num Members
E: MEM ALLOC = MEM Allocations
F: NUMA STATS = Numa Statistics
* G: SIZE = MEM Size (MB)
* H: ACTV = MEM Active (MB)
I: MCTL = MEM Ctl (MB)
J: SWAP STATS = Swap Statistics (MB)
K: CPT = MEM Checkpoint (MB)
L: COW = MEM Cow (MB)
* M: OVHD = MEM Overhead (MB)
N: CMT = MEM Committed (MB)
O: RESP? = MEM Responsive?
ESX disk adapter
Current Field order: ABCDEfgHIjklm
* A: ADAPTR = Adapter Name
* B: LUNID = Lun Id
* C: WID = World Id
* D: NLUNS = Num Luns
* E: NVMS = Num VMs
F: SHARES = Shares
G: BLKSZ = Block Size (bytes)
* H: QSTATS = Queue Stats
* I: IOSTATS = I/O Stats
J: LATSTATS = Latency stats (ms)
K: ERRSTATS/s = Error Stats
L: PAESTATS/s = PAE Stats
M: SPLTSTATS/s = SPLIT Stats
ESX disk device
Current Field order: ABCdeFGhijk
* A: DEVICE = Device Name
* B: ID = Path/World/Partition Id
* C: NUM = Num of Objects
D: SHARES = Shares
E: BLKSZ = Block Size (bytes)
* F: QSTATS = Queue Stats
* G: IOSTATS = I/O Stats
H: LATSTATS = Latency stats (ms)
I: ERRSTATS/s = Error Stats
J: PAESTATS/s = PAE Stats
K: SPLTSTATS/s = SPLIT Stats
ESX disk VM
Current Field order: ABCDEfgHIjk
* A: ID = World Id
* B: GID = Grp Id
* C: NAME = Group/World Name
* D: DEVICE = Device Name
* E: NUM = Num of Objects
F: SHARES = Shares
G: BLKSZ = Block Size (bytes)
* H: QSTATS = Queue Stats
* I: IOSTATS = I/O Stats
J: LATSTATS = Latency stats (ms)
K: ERRSTATS/s = Error Stats
ESX nic
Current Field order: ABcDEFGHIJKLm
* A: PORT ID = Port Id
* B: UPLINK = Uplink(Y/N)
C: PNIC = Physical Nic Properties
* D: USED BY = Used By Name
* E: DTYP = Device Type
* F: DNAME = Device Name
* G: PKTTX/s = Packets Tx/s
* H: MbTX/s = MegaBits Tx/s
* I: PKTRX/s = Packets Rx/s
* J: MbRX/s = MegaBits Rx/s
* K: DRPTX/s = %Packets Dropped (Tx)
* L: DRPRX/s = %Packets Dropped (Rx)
M: ACTN/s = Actions/s
man esxtop
ESXTOP(1) VMware ESX Server Manual ESXTOP(1)
NAME
esxtop, resxtop - display ESX Server resource utilization statistics
COPYRIGHT
VMware ESX Server is Copyright 2006-2007 VMware, Inc. All rights
reserved.
SYNOPSIS
esxtop [-] [h] [v] [b] [s] [a] [c configfile] [R vm-support_dir_path]
[d delay] [n iter]
resxtop [-] [h] [v] [b] [s] [a] [c configfile] [R vm-support_dir_path]
[d delay] [n iter] [-server host-name] [-portnumber socket-port]
[-username user-name]
DESCRIPTION
esxtop provides a fine-grain look at resource utilization in real time.
For more information, please refer to the official documentation,
available at <http://www.vmware.com/info?id=193>. Esxtop runs on the
ESX Server Service Console. It can only be run by user root. It can be
run in three different modes; interactive (default), batch and replay.
Worlds in this document refer to ESX Server VMKernel schedulable enti-
ties, similar to processes in other operating systems.
resxtop provides a fine-grain look at resource utilization of a speci-
fied host in real time. Resxtop connects to the host using a user name
and password. Options --server, --portnumber and --username can be used
to specify the connection parameters. The resxtop user will be prompted
to input the password. Resxtop can be run in three different modes in
the same way as esxtop; interactive (default), batch and replay.
Interactive Mode
This is the default mode for esxtop. This mode displays statistics
under four broad categories. Each category is displayed on a seperate
screen. There are screens displaying CPU, memory, storage and network
resource usage. A help menu is available for each of the four screens.
In this mode there are several command line options available.
Command-line Options
h Prints help for esxtop command-line options.
v Prints esxtop version number.
s Tells esxtop to run in secure mode. This disables the interactive
command to change delay between screen updates.
a Show all statistics. This option overides ~/.esxtop310rc configu-
ration file setups and shows all statistics.
c Load user-defined configuration file. Default configuration file
is ~/.esxtop310rc. User can specify the filename to be used as
configuration file. User-defined configuration file will be
default file name for saving configuration with W command.
d Specifies the delay between screen updates. Default is 5 seconds.
Minimum is 2 seconds. You can change this with the interactive
command 's'. If a delay of less than 2 seconds is specified then
the delay is set to 2 seconds.
n Number of iterations. Update the display 'n' number of times and
then exit.
Common Statistics Description
Several statistics are displayed on the different screens while esxtop
is running in interactive mode. Statistics listed in this section are
common across all four screens.
uptime
This first line found at the top of each of the four esxtop
screens displays the current time, time since last reboot and num-
ber of currently running Worlds.
On this first line CPU load averages are also displayed for the
CPU, storage and network screens. The load averages over the past
1, 5 and 15 minutes are displayed. Load averages take into account
both running and ready-to-run Worlds. A load average of 1.00
implies that all the physical CPUs are fully accounted for. A load
average of 2.00 implies that the ESX Server may be in need of
twice as many physical CPUs as currently available.
For the memory screen memory overcommit averages are also dis-
played on the first line. The memory overcommit averages over the
past 1, 5 and 15 minutes are displayed. A memory overcommit aver-
age of 0.37 implies that the memory is overcommit by 37%.
Common Interactive Commands
Several single-key commands are recognized while esxtop is running in
interactive mode. Commands listed in this section are common across all
four screens. The command to specify the delay between updates is dis-
abled if the s option has been given on the command line. All sorting
interactive commands sort in descending order.
h or ?
Displays a help menu for the current screen giving a brief summary
of commands, and the status of secure mode.
space
Immediately updates the current screen.
^L Erases and redraws the current screen.
f or F
Displays a screen to add or remove statistics columns (fields) to
or from the current screen. See below for more information.
o or O
Displays a screen to change order of statistics columns on the
current screen. See below for more information.
# Change the number of statistics rows to show. You will be prompted
to enter the number. Any value greater than 0 overrides automatic
determination of the number of rows to show, which is based on
window size measurement. This number updated on any one of the
four esxtop screens will be used for all esxtop screens.
s Change the delay between updates. You will be prompted to enter
the delay time, in seconds, between updates. Fractional values are
recognized down to microseconds. The default value is 5 seconds.
The minimum value is 2 seconds. This command is not available in
secure mode.
W Write current setup to ~/.esxtop310rc, or to user-defined configu-
ration file specified by -c option. This is the recommended way
to write an esxtop configuration file. See below for more informa-
tion.
q Quit.
c switch to CPU resource utilization screen.
m switch to memory resource utilization screen.
d switch to storage (disk) adapter resource utilization screen.
u switch to storage (disk) device resource utilization screen.
v switch to storage (disk) VM resource utilization screen.
n switch to network resource utilization screen.
The Statistics Columns (fields) and Order Screens
After pressing f, F, o or O you will be shown a screen specifying the
field order on the top line and short descriptions of the field con-
tents. The field order string uses the following syntax: if the letter
in the field string corresponding to a field is upper case, the field
will be displayed. This is also indicated by an asterisk in front of
the field description. The order of the fields corresponds to the
order of the letters in the string.
From the field select screen you can toggle the display of a field by
pressing the corresponding letter.
From the order screen you may move a field to the left by pressing the
corresponding upper case letter and to the right by pressing the lower
case one.
Configuration File
esxtop reads its default configuration from ~/.esxtop310rc or from
specified file name with -c option. This configuration file contains
five lines. The first four lines contain lower and upper case letters
to specify which fields in what order are to be displayed on the CPU,
memory, storage and network screens. The letters correspond to the let-
ters in the Fields or Order screens for the respective esxtop screens.
Editing this file by hand is not recommended. Select fields and order
in a running esxtop process and save this file using the W interactive
command. The fifth line contains information on the other options.
Most important, if you have saved a configuration in secure mode, you
will not get an insecure esxtop without removing the lower 's' from
this fifth line of your ~/.esxtop310rc. A digit specifies the delay
time between updates. As in interactive mode, a lower 'c', 'm', 'd'
and 'n' determines the screen esxtop starts on.
CPU Screen
This screen displays server-wide and per Resource Pool/World CPU uti-
lization statistics. By default, World CPU utilization statistics are
aggregated into Resource Pools of Worlds. For Worlds belonging to a
Virtual Machine, these Resource Pools correspond to running Virtual
Machines. All other Worlds are logically aggregated into Resource Pools
to save vertical window space.
Statistics Descriptions
PCPU(%)
This line displays the percentage CPU utilization per physical CPU
and the total average physical CPU utilization.
LCPU(%)
This line displays the percentage CPU utilization per logical CPU.
The CPU used percentages for the logical CPUs belonging to a pack-
age add up to 100%. This line is displayed only if hyper-thread-
ing is present and enabled.
CCPU(%)
This line displays the percentages of total CPU time as reported
by the ESX Server Service Console. us is for percentage user
time, sy is for percentage system time, id is for percentage idle
time and wa is for percentage wait time. cs/sec is for the con-
text switches per second recorded by the ESX Server Service
Console.
ID The Resource Pool ID of the running World's Resource Pool or World
ID of running World.
GID The Resource Pool ID of running World's Resource Pool.
NAME The name of running World's Resource Pool or name of running
World.
NWLD The number of worlds in running World's Resource Pool. If a
Resource Pool is expanded using the interactive command e (see
interactive commands below) then NWLD for all the resulting Worlds
belonging to the Resource Pool will be 1 (some Resource Pool like
the console Resource Pool have only 1 member).
%STATE TIMES
This is a set of CPU statistics made up of the following percent-
ages. Note that for a World, the percentages are a percentage of
one physical CPU.
%USED
The percentage physical CPU used by the Resource Pool/World.
%SYS The percentage of time spent in the ESX Server VMKernel on behalf
of the Resource Pool/World to process interrupts and to perform
other system activities. This time is part of used time used to
calculate %USED above.
%OVRLP
The Percentage of system time that was spent on behalf of some
other Resource Pool/World while Resource Pool/World was scheduled.
This time is not included in the %SYS for Resource Pool/World. For
example, if Resource Pool A was currently scheduled and a network
packet for Resource Pool B was being processed by the ESX Server
VMKernel, then the time spent doing so appears as %OVRLP for
Resource Pool A and %SYS for Resource Pool B.
%RUN Percentage of total time scheduled. This time does not account for
hyper-threading and system time. Hence, on a hyper-threading
enabled server, the %RUN can be twice as large as %USED.
%WAIT
The total percentage of time the Resource Pool/World spent in wait
state. Note that this percentage includes the percentage of time
the Resource Pool/World was idle.
%CSTP
The percentage of time the Resource Pool/World spent in ready, co-
deschedule state. This statistic is for internal to VMware use
only.
%IDLE
The percentage of time the Resource Pool/World was idle. Subtract-
ing this percentage from %WAIT above gives you the percentage of
time the Resource Pool/World was waiting on some event.
%RDY The percentage of time the Resource Pool/World was ready to run.
%MLMTD
Percentage of time the ESX Server VMKernel deliberately didn't run
the Resource Pool/World because that would violate the Resource
Pool/World's limit setting. Even though the Resource Pool/World is
ready to run when it is prevented from running in this way, the
%MLMTD time is not included in %RDY time.
EVENT COUNTS/s
This is a set of CPU statistics made up of the following, per sec-
ond, event rates. These statistics are for internal to VMware use
only.
CPU ALLOC
This is a set of CPU statistics made up of the following CPU allo-
cation configuration parameters.
AMIN Resource Pool/World attribute reservation.
AMAX Resource Pool/World attribute limit.
ASHRS
Resource pool attribute shares.
SUMMARY STATS
This is a set of CPU statistics made up of the following CPU con-
figuration parameters and statistics. Note that these statistics
are only applicable to Worlds and not Resource Pools.
AFFINITY
Bit mask showing the current scheduling affinity for the World.
HTSHARING
current hyper-threading configuration.
CPU The physical or logical processor the World was found to be run-
ning on.
HTQ World is currently quaratined or not. 'N' implies no and 'Y'
implies yes.
TIMER/s
The timer rate the World is currently requesting for.
Interactive Commands
e Expand/Rollup Resource Pool/Worlds CPU statistics. Allows viewing
CPU resource utilization statistics broken down by individual
Worlds belonging to a Resource Pool.
U Sort Resource Pools/Worlds by Resource Pool %USED column.
R Sort Resource Pools/Worlds by Resource Pool %RDY column.
N Sort Resource Pools/Worlds by GID column. This is the default sort
order.
V Display VM instances only.
Memory Screen
This screen displays server-wide and per Resource Pool memory utiliza-
tion statistics. As on the CPU screen, Resource Pools correspond to
running Virtual Machines or Worlds that consume memory. Note the dis-
tinction between machine memory and physical memory in this section.
Statistics Descriptions
PMEM (MB)
This line displays the machine memory statistics for the server.
All numbers are in megabytes. total is for the total amount of
machine memory in the server, cos is for the amount of machine
memory allocated to the ESX Server Service Console, vmk is for the
amount of machine memory being used by the ESX Server VMKernel,
other is for the amount of machine memory being used by everything
other than the ESX Server Service Console and ESX Server VMKernel
and free is for the amount of machine memory that is free.
VMKMEM (MB)
This line displays the machine memory statistics for the ESX
Server VMKernel. All numbers are in megabytes. managed is for
the total amount of machine memory managed by the ESX Server VMK-
ernel, minfree is for the minimum amount of machine memory that
the ESX Server VMKernel would like to keep free, rsvd is for the
amount of machine memory that is currently reserved, ursvd is for
the amount of machine memory that is currently unreserved and
state is for the memory state as reported by the ESX Server VMKer-
nel. Possible values are high, soft, hard and low. Here, high
implies that the machine memory is not under any pressure and low
implies that the machine memory is under pressure.
COSMEM (MB)
This line displays the memory statistics as reported by the ESX
Server Service Console. All numbers are in megabytes. free is
for the amount of idle machine memory, swap_t is for the total
swap configured, swap_f is for the amount of swap free, r/s is for
the rate at which memory is swapped in from disk and w/s is for
the rate at which memory is swapped to disk.
NUMA (MB)
This line displays the ESX Server NUMA statistics. This line is
only displayed if the ESX Server is running on a NUMA server. All
numbers are in megabytes. For each NUMA node in the server there
are two statistics that are displayed. The first statistic is the
total amount of machine memory in the NUMA node that is managed by
the ESX Server. The second statistic, that is displayed within
round brackets, is the amount of machine memory in the node that
is currently free.
PSHARE (MB)
This line displays the ESX Server page-sharing statistics. All
numbers are in megabytes. shared is for the amount of physical
memory that is being shared, common is for the amount of machine
memory that is common across World(s) and saving is for the amount
of machine memory that is saved due to page-sharing.
SWAP (MB)
This line displays the ESX Server swap usage statistics. All num-
bers are in megabytes. curr is for the current swap usage, target
is for where the ESX Server expects the swap usage to be, r/s is
for the rate at which memory is swapped in by the ESX Server from
disk and w/s is for the rate at which memory is swapped to disk by
the ESX Server.
MEMCTL (MB)
This line displays the Memory balloon statistics. All numbers are
in megabytes. curr is for the total amount of physical memory
reclaimed using the vmmemctl modules, target is for the total
amount of physical memory the ESX Server would like to reclaim
using the vmmemctl modules and max is for the maximum amount of
physical memory the ESX Server can reclaim using the vmmemctl mod-
ules.
AMIN Resource pool attribute reservation.
AMAX Resource pool attribute limit.
ASHRS
Resource pool attribute shares.
NHN Current Home Node for Resource Pool. This statistic is only appli-
cable on NUMA systems.
NRMEM (MB)
Current amount of remote memory being accessed by Resource Pool.
This statistic is only applicable on NUMA systems.
NLMEM (MB)
Current amount of local memory being accessed by Resource Pool.
This statistic is only applicable on NUMA systems.
N%L Current percentage memory being accessed by Resource Pool that is
local. This statistic is only applicable on NUMA systems.
GST_NDx (MB)
The guest memory being allocated for Resource Pool on NUMA node x.
"x" is the node number. This statistic is only applicable on NUMA
systems.
OVD_NDx (MB)
The VMM overhead memory being allocated for Resource Pool on NUMA
node x. "x" is the node number. This statistic is only applica-
ble on NUMA systems.
MEMSZ (MB)
The amount of physical memory allocated to a Resource Pool.
SZTGT (MB)
The amount of machine memory the ESX Server VMKernel wants to
allocate to Resource Pool.
TCHD (MB)
The working set estimate for Resource Pool.
%ACTV
Percentage of guest physical memory that is being referenced by
the guest. This is an instantaneous value.
%ACTVS
Percentage of guest physical memory that is being referenced by
the guest. This is a slow moving average.
%ACTVF
Percentage of guest physical memory that is being referenced by
the guest. This is a fast moving average.
%ACTVN
Percentage of guest physical memory that will be referenced by the
guest. This is an estimation. This statistic is for internal to
VMware use only.
MCTL?
memory balloon driver is installed or not. 'N' implies no and 'Y'
imples yes.
MCTLSZ (MB)
The amount of physical memory reclaimed from Resource Pool by way
of ballooning.
MCTLTGT (MB)
The amount of physical memory the ESX Server would like to reclaim
from Resource Pool by way of ballooning.
MCTLMAX (MB)
The maximum amount of physical memory the ESX Server can reclaim
from Resource Pool by way of ballooning. This maximum is depen-
dent on guest operating system type.
SWCUR (MB)
Current swap usage by Resource Pool.
SWTGT (MB)
Where the ESX Server expectd the Swap usage by the Resource Pool
to be.
SWR/s (MB)
Rate at which memory is being swapped in by the ESX Server from
disk for the Resource Pool.
SWW/s (MB)
Rate at which Resource Pool memory is being swapped to disk by the
ESX Server.
CPTRD (MB)
Amount of data read from checkpoint file.
CPTTGT (MB)
Size of checkpoint file.
ZERO (MB)
Resource Pool physical pages that are zeroed.
SHRD (MB)
Resource Pool physical pages that are shared.
SHRDSVD (MB)
Machine pages that are saved due to Resource Pool shared pages.
COWH (MB)
This statistic is for internal to VMware use only.
OVHDUW (MB)
Current space overhead for the user world. This statistic is for
internal to VMware use only.
OVHD (MB)
Current space overhead for Resource Pool.
OVHDMAX (MB)
Maximum space overhead that may be incurred by Resource Pool.
MEM Committed (MB)
This is a set of memory statistics made up of minimum memory com-
mitment target, current memory commitment target, charged memory
commitment, and pages per share. These statistics are for internal
to VMware use only.
MEM Responsive
This statistic is for internal to VMware use only.
Interactive Commands
V Display VM instances only.
M Sort Resource Pools by Resource Pool MEMSZ column.
B Sort Resource Pools by Resource Pool MCTLSZ column.
N Sort Resource Pools by GID column. This is the default sort order.
Storage Adapter Screen
This screen displays server-wide storage utilization statistics. On
this screen statistics are aggregated per storage adapter by default.
Statistics can also be viewed per storage channel, target, path or
world using a LUN. See below for more information.
Statistics Descriptions
ADAPTR
The name of the storage adapter.
CID The storage adapter channel id. This id is only visible if the
corresponding adapter is expanded. See interactive command 'e'
below.
TID The storage adapter channel target id. This id is only visible if
the corresponding adapter and channel are expanded. See interac-
tive commands 'e' and 'a' below.
LID The storage adapter channel target LUN id. This id is only visible
if the corresponding adapter, channel and target are expanded. See
interactive commands 'e', 'a' and 't' below. ADAPTR:CID:TID:LID
specifies one path. When the statistics are expanded to lun
level, esxtop shows the path statistics.
WID The storage adapter channel target LUN World id. This id is only
visible if the corresponding adapter, channel, target and LUN are
expanded. See interactive commands 'e', 'a', 't' and 'l' below.
The world statistics are per world per device.
NCHNS
The number of channels.
NTGTS
The number of targets.
NLUNS
The number of LUNs.
NVMS The number of Worlds.
SHARES
The number of shares. This statistic is only applicable to
worlds.
BLKSZ
The block size in bytes. This statistic is only applicable to
LUNs.
AQLEN
The storage adapter queue depth. This is the maximum number of
ESX Server VMKernel active commands that the adapter driver is
configured to support.
LQLEN
The LUN queue depth. This is the maximum number of ESX Server
VMKernel active commands that the LUN is allowed to have.
WQLEN
The World queue depth. This is the maximum number of ESX Server
VMKernel active commands that the World is allowed to have. Note
that this is a per LUN maximum for the World.
ACTV The number of commands in the ESX Server VMKernel that are cur-
rently active. This statistic is only applicable to worlds and
LUNs.
QUED The number of commands in the ESX Server VMKernel that are cur-
rently queued. This statistic is only applicable to worlds and
LUNs.
%USD The percentage of queue depth used by ESX Server VMKernel active
commands. This statistic is only applicable to worlds and LUNs.
LOAD The ratio of ESX Server VMKernel active commands plus ESX Server
VMKernel queued commands to queue depth. This statistic is only
applicable to worlds and LUNs.
CMDS/s
The number of commands issued per second. If multiple paths are
connected to the same lun, this statistic is per path when it is
expanded to luns; it is per world per device, when it is expanded
to worlds.
READS/s
The number of read commands issued per second. If multiple paths
are connected to the same lun, this statistic is per path when it
is expanded to luns; it is per world per device, when it is
expanded to worlds.
WRITES/s
The number of write commands issued per second. If multiple paths
are connected to the same lun, this statistic is per path when it
is expanded to luns; it is per world per device, when it is
expanded to worlds.
MBREAD/s
The megabytes read per second. If multiple paths are connected to
the same lun, this statistic is per path when it is expanded to
luns; it is per world per device, when it is expanded to worlds.
MBWRTN/s
The megabytes written per second. If multiple paths are connected
to the same lun, this statistic is per path when it is expanded to
luns; it is per world per device, when it is expanded to worlds.
DAVG/cmd
The average device latency (millisecs) per command.
KAVG/cmd
The average ESX Server VMKernel latency (millisecs) per command.
GAVG/cmd
The average Guest OS latency (millisecs) per command.
QAVG/cmd
The average queue latency (millisecs) per command.
DAVG/rd
The average device read latency (millisecs) per read.
KAVG/rd
The average ESX Server VMKernel read latency (millisecs) per read.
GAVG/rd
The average Guest OS read latency (millisecs) per read.
QAVG/rd
The average queue read latency (millisecs) per read.
DAVG/wr
The average device write latency (millisecs) per write.
KAVG/wr
The average ESX Server VMKernel write latency (millisecs) per
write.
GAVG/wr
The average Guest OS write latency (millisecs) per write.
QAVG/wr
The average queue write latency (millisecs) per write.
ABRTS/s
The number of commands aborted per second. If multiple paths are
connected to the same lun, this statistic is per path when it is
expanded to luns; it is per world per device, when it is expanded
to worlds.
RESETS/s
The number of commands reset per second. If multiple paths are
connected to the same lun, this statistic is per path when it is
expanded to luns; it is per world per device, when it is expanded
to worlds.
PAECMD/s
The number of PAE commands per second. This statistic is not
applicable to worlds.
PAECP/s
The number of PAE copies per second. This statistic is not appli-
cable to worlds.
SPLTCMD/s
The number of split commands per second. This statistic is not
applicable to worlds.
SPLTCP/s
The number of split copies per second. This statistic is not
applicable to worlds.
Storage Device Screen
This screen displays server-wide storage utilization statistics. On
this screen statistics are aggregated per storage device by default.
Statistics can also be viewed per path, world, or partition. See below
for more information.
Statistics Descriptions
DEVICE
The name of the storage device.
PATH The path name. This name is only visible if the corresponding
device is expanded to paths. See interactive command 'p' below.
WORLD
The world id. This id is only visible if the corresponding device
is expanded to worlds. See interactive command 'e' below. The
world statistics are per world per device.
PARTITION
The partition id. This id is only visible if the corresponding
device is expanded to partitions. See interactive command 't'
below.
NPH The number of paths.
NWD The number of worlds.
NPN The number of partitions.
SHARES
The number of shares. This statistic is only applicable to
worlds.
BLKSZ
The block size in bytes.
NUMBLKS
The number of blocks of the device.
DQLEN
The storage device queue depth. This is the maximum number of ESX
Server VMKernel active commands that the device is configured to
support.
WQLEN
The World queue depth. This is the maximum number of ESX Server
VMKernel active commands that the World is allowed to have. Note
that this is a per device maximum for the World. It is valid only
if the corresponding device is expanded to worlds.
ACTV The number of commands in the ESX Server VMKernel that are cur-
rently active. This statistic is only applicable to worlds and
devices.
QUED The number of commands in the ESX Server VMKernel that are cur-
rently queued. This statistic is only applicable to worlds and
devices.
%USD The percentage of queue depth used by ESX Server VMKernel active
commands. This statistic is only applicable to worlds and
devices.
LOAD The ratio of ESX Server VMKernel active commands plus ESX Server
VMKernel queued commands to queue depth. This statistic is only
applicable to worlds and devices.
CMDS/s
The number of commands issued per second.
READS/s
The number of read commands issued per second.
WRITES/s
The number of write commands issued per second.
MBREAD/s
The megabytes read per second.
MBWRTN/s
The megabytes written per second.
DAVG/cmd
The average device latency (millisecs) per command.
KAVG/cmd
The average ESX Server VMKernel latency (millisecs) per command.
GAVG/cmd
The average Guest OS latency (millisecs) per command.
QAVG/cmd
The average queue latency (millisecs) per command.
DAVG/rd
The average device read latency (millisecs) per read.
KAVG/rd
The average ESX Server VMKernel read latency (millisecs) per read.
GAVG/rd
The average Guest OS read latency (millisecs) per read.
QAVG/rd
The average queue read latency (millisecs) per read.
DAVG/wr
The average device write latency (millisecs) per write.
KAVG/wr
The average ESX Server VMKernel write latency (millisecs) per
write.
GAVG/wr
The average Guest OS write latency (millisecs) per write.
QAVG/wr
The average queue write latency (millisecs) per write.
ABRTS/s
The number of commands aborted per second.
RESETS/s
The number of commands reset per second.
PAECMD/s
The number of PAE commands per second. This statistic is only
applicable to paths.
PAECP/s
The number of PAE copies per second. This statistic is only
applicable to paths.
SPLTCMD/s
The number of split commands per second. This statistic is only
applicable to paths.
SPLTCP/s
The number of split copies per second. This statistic is only
applicable to paths.
Interactive Commands
e Expand/Rollup storage world statistics. Allows viewing storage
resource utilization statistics broken down by individual worlds
belonging to an expanded storage device. You will be prompted to
enter the device name. The statistics are per world per device.
p Expand/Rollup storage path statistics. Allows viewing storage
resource utilization statistics broken down by individual paths
belonging to an expanded storage device. You will be prompted to
enter the device name.
t Expand/Rollup storage partition statistics. Allows viewing storage
resource utilization statistics broken down by individual parti-
tions belonging to an expanded storage device. You will be
prompted to enter the device name.
r Sort by READS/s column.
w Sort by WRITES/s column.
R Sort by MBREAD/s column.
T Sort by MBWRTN/s column.
N Sort first by DEVICE column, then by PATH/WORLD/PARTITION column.
This is the default sort order.
Storage VM Screen
This screen displays VM-centric storage statistics. On this screen,
statistics are aggregated on a per-resource-pool basis by default. One
VM has one corresponding resource pool, so, they are equivalent to per-
VM statistics. Statistics can also be viewed on a per-world, or a per-
world-per-device basis.
Statistics Descriptions
ID The Resource Pool ID of the running World's Resource Pool or ,B
World ID of runnng World.
GID The Resource Pool ID of running World's Resource Pool.
NAME The name of running World's Resource Pool or name of the running
World.
Device
The storage device name. This name is only visible if correspond-
ing world is expanded to devices. See interactive command 'i'
below.
NWD The number of worlds.
NDV The number of devices. It is valid only if the corresponding
resource pool is expanded to worlds
SHARES
The number of shares. This statistic is only applicable to
worlds. It is valid only if the corresponding resource pool is
expaneded to worlds
BLKSZ
The block size in bytes. It is valid only if the corresponding
world is expanded to devices.
NUMBLKS
The number of blocks of the device. It is valid only if the corre-
sponding world is expanded to devices.
DQLEN
The storage device queue depth. This is the maximum number of ESX
Server VMKernel active commands that the device is configured to
support. It is valid only if the corresponding world is expanded
to devices.
WQLEN
The World queue depth. This is the maximum number of ESX Server
VMKernel active commands that the World is allowed to have. It is
valid only if the corresponding world is expanded to worlds.
ACTV The number of commands in the ESX Server VMKernel that are cur-
rently active. This statistic is only applicable to worlds and
devices.
QUED The number of commands in the ESX Server VMKernel that are cur-
rently queued. This statistic is only applicable to worlds and
devices.
%USD The percentage of queue depth used by ESX Server VMKernel active
commands. This statistic is only applicable to worlds and
devices.
LOAD The ratio of ESX Server VMKernel active commands plus ESX Server
VMKernel queued commands to queue depth. This statistic is only
applicable to worlds and devices.
CMDS/s
The number of commands issued per second.
READS/s
The number of read commands issued per second.
WRITES/s
The number of write commands issued per second.
MBREAD/s
The megabytes read per second.
MBWRTN/s
The megabytes written per second.
DAVG/cmd
The average device latency (millisecs) per command.
KAVG/cmd
The average ESX Server VMKernel latency (millisecs) per command.
GAVG/cmd
The average Guest OS latency (millisecs) per command.
QAVG/cmd
The average queue latency (millisecs) per command.
DAVG/rd
The average device read latency (millisecs) per read.
KAVG/rd
The average ESX Server VMKernel read latency (millisecs) per read.
GAVG/rd
The average Guest OS read latency (millisecs) per read.
QAVG/rd
The average queue read latency (millisecs) per read.
DAVG/wr
The average device write latency (millisecs) per write.
KAVG/wr
The average ESX Server VMKernel write latency (millisecs) per
write.
GAVG/wr
The average Guest OS write latency (millisecs) per write.
QAVG/wr
The average queue write latency (millisecs) per write.
ABRTS/s
The number of commands aborted per second.
RESETS/s
The number of commands reset per second.
Interactive Commands
e Expand/Rollup storage world statistics. Allows viewing storage
resource utilization statistics broken down by individual worlds
belonging to a group. You will be prompted to enter the group id.
The statistics are per world.
l Expand/Rollup storage device(=lun) statistics. Allows viewing
storage resource utilization statistics broken down by individual
devices belonging to an expanded world. You will be prompted to
enter the world id.
V Display VM instances only.
r Sort by READS/s column.
w Sort by WRITES/s column.
R Sort by MBREAD/s column.
T Sort by MBWRTN/s column.
N Sort first by VM column, then by WORLD column. This is the
default sort order.
Network Screen
This screen displays server-wide network utilization statistics. On
this screen statistics are arranged per port per configured virtual
network device. For physical NIC statistics, see the row corresponding
to the port that the physical NIC is connected to. For virtual NIC con-
figured in a particular Virtual Machine statistics, see the row corre-
sponding to the port that the virtual NIC is connected to.
Statistics Descriptions
PORT ID
The virtual network device port id.
UPLINK
Y implies the corresponding port is an uplink. N implies it is
not.
UP Y implies the corresponding link is up. N implies it is not.
SPEED
The link speed in MegaBits per second.
FDUPLX
Y implies the corresponding link is operating at full duplex. N
implies it is not.
USED BY
The virtual network device port user.
DTYP The virtual network device type. H implies HUB and S implies
switch.
DNAME
The virtual network device name.
PKTTX/s
The number of packets transmitted per second.
PKTRX/s
The number of packets received per second.
MbTX/s
The MegaBits transmitted per second.
MbRX/s
The MegaBits received per second.
%DRPTX
The percentage of transmit packets dropped.
%DRPRX
The percentage of receive packets dropped.
ACTN/s
Number of actions per second. This statistic is for internal to
VMware use only.
Interactive Commands
T Sort by MbTX/s column.
R Sort by MbRX/s column.
t Sort by PKTTX/s column.
r Sort by PKTRX/s column.
N Sort by PORT ID column. This is the default sort order.
Batch Mode
Allows collecting and saving resource utilization statistics in a file.
Running in this mode is a two step process.
Step one requires running esxtop in interactive mode, switching to
each of the four available screens, selecting the columns on each
screen you are interested in and saving this configuration in the
~/.esxtop310rc file using the W interactive command. See Interactive
Mode above. Note that the global stats are always available in batch
mode.
Step two requires running esxtop in batch mode and redirecting the
output to a file. This may be done as follows: esxtop -b > foo.csv.
Note that the file name needs to have a .csv extension. Esxtop does not
enforce this, but post-processing tools mentioned next require this.
Statistics collected in batch mode may be post-processed using such MS
Windows applications like Excel and Perfmon.
In this mode, esxtop will not accept interactive commands. Esxtop in
batch mode runs until it produces the number of iterations requested
(see command-line option n below for more details) or until killed
using CTRL c.
In this mode there are several command line options available.
Command-line Options
b Run esxtop in Batch mode.
d Specifies the delay between statistics snapshots. Default is 5
seconds. Minimum is 2 seconds. If a delay of less than 2 seconds
is specified then the delay is set to 2 seconds.
n Number of iterations. Collect and save statistics this number of
times and then exit.
Replay Mode
Replays resource utilization statistics collected using vm-support (see
vm-support man page for more information). Running in this mode is a
two step process.
Step one requires running vm-support in snapshot mode on the ESX
Server Service Console. This may be done as follows: vm-support -S -d
duration -i interval. The resulting zipped tar file needs to be un-
zipped and un-tar'ed before use by esxtop replay mode.
Step two requires running esxtop in replay mode. This may be done as
follows: esxtop -R vm-support_dir_path. Note that it is not required
to run esxtop replay mode on the ESX Server Service Console.
Replay mode can be run to produce Batch Mode style output (see com-
mand-line option b below for more information).
In replay mode, esxtop accepts the same set of interactive commands as
in Interactive Mode. Esxtop in replay mode runs until there are no
more vm-support collected snapshots to be read or esxtop produces the
number of iterations requested (see command-line option n below for
more details).
In this mode there are several command line options available.
Command-line Options
R Specifies the path to the vm-support collected snapshot's direc-
tory.
b Run esxtop in Batch mode.
d Specifies the delay between screen updates. Default is 5 seconds.
Minimum is 2 seconds. If a delay of less than 2 seconds is speci-
fied then the delay is set to 2 seconds.
n Number of iterations. Update the display this number of times and
then exit.
FILES
~/.esxtop310rc The personal configuration file.
SEE ALSO
vm-support(1), cpu(8), mem(8), diskbw(8), numa(8)
DISCLAIMER
Based on top source code from the OpenBSD distribution.
Copyright (c) 1997, Jason Downs. All rights reserved.
Copyright (c) 1984, 1989, William LeFebvre, Rice University
Copyright (c) 1989, 1990, 1992, William LeFebvre,
Northwestern University
Redistribution and use in source and binary forms, with or without mod-
ification, are permitted provided that the following conditions are
met:
1. Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
2. Redistributions in binary form must reproduce the above copyright
notice, this list of conditions and the following disclaimer in the
documentation and/or other materials provided with the distribution.
THIS SOFTWARE IS PROVIDED BY THE AUTHOR(S) ``AS IS'' AND ANY EXPRESS OR
IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WAR-
RANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
DISCLAIMED. IN NO EVENT SHALL THE AUTHOR(S) BE LIABLE FOR ANY DIRECT,
INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
(INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SER-
VICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIA-
BILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
SUCH DAMAGE.
ESX 3.5 June 7, 2007 ESXTOP(1)