podman-stats - Man Page

Display a live stream of one or more container's resource usage statistics


podman stats [options] [container]

podman container stats [options] [container]


Display a live stream of one or more containers' resource usage statistics

Note:  Podman stats does not work in rootless environments that use CGroups V1. Podman stats relies on CGroup information for statistics, and CGroup v1 is not supported for rootless use cases.

Note: Rootless environments that use CGroups V2 are not able to report statistics about their networking usage.


--all, -a

Show all containers.  Only running containers are shown by default


Pretty-print container statistics to JSON or using a Go template

Valid placeholders for the Go template are listed below:

.AvgCPUAverage CPU, full precision float
.AVGCPUAverage CPU, formatted as a percent
.BlockInputTotal data read from block device
.BlockIOTotal data read/total data written to block device
.BlockOutputTotal data written to block device
.ContainerIDContainer ID, full (untruncated) hash
.ContainerStats ...Nested structure, for experts only
.CPUPercent CPU, full precision float
.CPUNanoCPU Usage, total, in nanoseconds
.CPUPercPercentage of CPU used
.CPUSystemNanoCPU Usage, kernel, in nanoseconds
.DurationSame as CPUNano
.IDContainer ID, truncated
.MemLimitMemory limit, in bytes
.MemPercMemory percentage used
.MemUsageMemory usage
.MemUsageBytesMemory usage (IEC)
.NameContainer Name
.NetIONetwork IO
.NetworkNetwork I/O, separated by network interface
.PerCPUCPU time consumed by all tasks [1]
.PIDsNumber of PIDs
.PIDSNumber of PIDs (yes, we know this is a dup)
.SystemNanoCurrent system datetime, nanoseconds since epoch
.UpDuration (CPUNano), in human-readable form
.UpTimeSame as Up

[1] Cgroups V1 only

When using a Go template, precede the format with table to print headers.

--interval, -i=seconds

Time in seconds between stats reports, defaults to 5 seconds.

--latest, -l

Instead of providing the container name or ID, use the last created container. Note: the last started container can be from other users of Podman on the host machine. (This option is not available with the remote Podman client, including Mac and Windows (excluding WSL2) machines)


Do not clear the terminal/screen in between reporting intervals


Disable streaming stats and only pull the first result, default setting is false


Do not truncate output


# podman stats -a --no-stream
ID             NAME              CPU %   MEM USAGE / LIMIT   MEM %   NET IO    BLOCK IO   PIDS
a9f807ffaacd   frosty_hodgkin    --      3.092MB / 16.7GB    0.02%   -- / --   -- / --    2
3b33001239ee   sleepy_stallman   --      -- / --             --      -- / --   -- / --    --
# podman stats --no-stream a9f80
ID             NAME             CPU %   MEM USAGE / LIMIT   MEM %   NET IO    BLOCK IO   PIDS
a9f807ffaacd   frosty_hodgkin   --      3.092MB / 16.7GB    0.02%   -- / --   -- / --    2
$ podman stats --no-trunc 3667 --format 'table {{ .ID }} {{ .MemUsage }}'
ID                                                                MEM USAGE / LIMIT
3667c6aacb06aac2eaffce914c01736420023d56ef9b0f4cfe58b6d6a78b7503  49.15kB / 67.17GB
# podman stats --no-stream --format=json a9f80
	"id": "a9f807ffaacd",
	"name": "frosty_hodgkin",
	"cpu_percent": "--",
	"mem_usage": "3.092MB / 16.7GB",
	"mem_percent": "0.02%",
	"netio": "-- / --",
	"blocki": "-- / --",
	"pids": "2"
# podman stats --no-stream --format "table {{.ID}} {{.Name}} {{.MemUsage}}" 6eae
ID             NAME           MEM USAGE / LIMIT
6eae9e25a564   clever_bassi   3.031MB / 16.7GB

Note: When using a slirp4netns network with the rootlesskit port handler, the traffic sent via the port forwarding is accounted to the lo device.  Traffic accounted to lo is not accounted in the stats output.

See Also



July 2017, Originally compiled by Ryan Cole rycole@redhat.com ⟨mailto:rycole@redhat.com⟩

Referenced By

podman(1), podman-container(1).

The man page podman-container-stats(1) is an alias of podman-stats(1).