Shulou(Shulou.com)06/01 Report--

This article shares with you the content of a sample analysis of the source code of the docker stats command. The editor thinks it is very practical, so share it with you as a reference and follow the editor to have a look.

This article is based on the source code of docker version 1.10.3 to analyze the source code of the docker stats command to see how the output data of the docker stats command is calculated from cgroups fs.

$docker stats nginx-test | CONTAINER | CPU% | MEM USAGE / LIMIT | MEM% | NET kB O | BLOCK Icano | |-- |-- | nginx-test | 0.005% | 4.268 MB / 1.041 GB | 0.41% | 1.296 kB / 648B | 7.463 MB / 0B |

Docker client related code entry can be referred to: / docker/docker/api/client/stats.go#141 docker daemon related code entry can be referred to: / docker/docker/daemon/daemon.go#1474

# # Source code analysis result # Cpu data: docker daemon records the value of this read / sys/fs/cgroup/cpuacct/docker/ [containerId] / cpuacct.usage as cpu_total_usage; and records the last read value as pre_cpu_total_usage

Read the cpu field value in / proc/stat and accumulate it to get system_usage; and record the last value as pre_system_usage

Read the records in / sys/fs/cgroup/cpuacct/docker/ [containerId] / cpuacct.usage_percpu to form an array per_cpu_usage_array

Docker stats's algorithm for calculating Cpu Percent:

Cpu_delta = cpu_total_usage-pre_cpu_total_usage; system_delta = system_usage-pre_system_usage; CPU% = ((cpu_delta / system_delta) * length (per_cpu_usage_array)) * 100.0

# Memory data:

Read the value of / sys/fs/cgroup/memory/docker/ [containerId] / memory.usage_in_bytes as mem_usage

If the container has limited memory, read / sys/fs/cgroup/memory/docker/ [id] / memory.limit_in_bytes as mem_limit, otherwise mem_limit = machine_mem

Docker stats's algorithm for calculating Memory data:

MEM USAGE = mem_usage MEM LIMIT = mem_limit MEM% = (mem_usage / mem_limit) * 100.0

Networt Stats data:

Get the veth* virtual network card EthInterface array corresponding to the container network namespace veth pairs in the host, and then loop each network card device in the array, read / sys/class/net//statistics/rx_bytes to get rx_bytes, and read / sys/class/net//statistics/tx_bytes to get the corresponding tx_bytes.

Add up the rx_bytes of all these virtual network cards to get the rx_bytes of the container.

Add up the tx_bytes of all these virtual network cards to get the tx_bytes of the container.

Docker stats's algorithm for calculating Network IO data:

NET I = rx_bytes NET O = tx_bytes

# Blkio Stats data: get the IoServiceBytesRecursive data of each block device: first read whether there is a valid value in / sys/fs/cgroup/blkio/docker/ [containerId] / blkio.io_serviced_recursive

If so, the value of / sys/fs/cgroup/blkio/docker/ [containerId] / blkio.io_service_bytes_recursive is returned.

If not, read the value return in / sys/fs/cgroup/blkio/docker/ [containerId] / blkio.throttle.io_service_bytes

Accumulate all the read field corresponding value in the IoServiceBytesRecursive data of each block device to get the blk_ read value of the container; accumulate all the write field corresponding value in the IoServiceBytesRecursive data of each block device to get the blk_ write value of the container

Docker stats's algorithm for calculating Block IO data:

BLOCK I = blk_read BLOCK O = blk_write

Thank you for reading! On the "docker stats command source code example analysis" this article is shared here, I hope the above content can be of some help to you, so that you can learn more knowledge, if you think the article is good, you can share it out for more people to see it!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.