首頁/ 汽車/ 正文

如何檢視多個pod的日誌流

如何檢視多個pod的日誌流

‬問題說明

Kubetail 是一個 bash 指令碼,它使您能夠將來自多個 pod 的日誌聚合(tail/follow)到一個流中

這與執行 `kubectl logs -f` 相同,但適用於多個 pod

‬使用舉例

直接tail應用

$ kubetail app2

指定‬應用‬和‬容器‬

$ kubetail app2 -c container1

可以‬使用‬-c tail多個‬容器‬

$ kubetail app2 -c container1 -c container2

使用‬逗號‬分隔‬多個‬應用‬

$ kubetail app1,app2

高階‬用法‬當然‬也不能‬少了‬正則‬表達‬式了‬

$ kubetail “^app1|。*my-demo。*” ——regex

‬指令碼分享

#!/bin/bashif [ -z “${KUBECTL_BIN}” ]; then if hash kubectl 2>/dev/null; then KUBECTL_BIN=‘kubectl’ elif hash kubectl。exe 2>/dev/null; then KUBECTL_BIN=‘kubectl。exe’ elif hash microk8s 2>/dev/null; then KUBECTL_BIN=‘microk8s。kubectl’ fifiif ! hash “${KUBECTL_BIN}” 2>/dev/null; then echo >&2 “kubectl is not installed” exit 1fireadonly PROGNAME=$(basename $0)calculate_default_namespace() { local config_namespace=$(${KUBECTL_BIN} config view ——minify ——output ‘jsonpath={。。namespace}’) echo “${KUBETAIL_NAMESPACE:-${config_namespace:-default}}”}# Sets default color ouput to ‘false’ if output is not a terminalterminal_aware_default_color=line[ ! -t 1 ] && terminal_aware_default_color=falsedefault_previous=“${KUBETAIL_PREVIOUS:-false}”default_since=“${KUBETAIL_SINCE:-10s}”default_namespace=$(calculate_default_namespace)default_follow=“${KUBETAIL_FOLLOW:-true}”default_prefix=“${KUBETAIL_PREFIX:-true}”default_line_buffered=“${KUBETAIL_LINE_BUFFERED:-}”default_colored_output=“${KUBETAIL_COLORED_OUTPUT:-$terminal_aware_default_color}”default_timestamps=“${KUBETAIL_TIMESTAMPS:-}”default_jq_selector=“${KUBETAIL_JQ_SELECTOR:-}”default_skip_colors=“${KUBETAIL_SKIP_COLORS:-7,8}”default_tail=“${KUBETAIL_TAIL:——1}”default_show_color_index=“${KUBETAIL_SHOW_COLOR_INDEX:-false}”namespace=“${default_namespace}”follow=“${default_follow}”prefix=“${default_prefix}”line_buffered=“${default_line_buffered}”colored_output=“${default_colored_output}”timestamps=“${default_timestamps}”jq_selector=“${default_jq_selector}”skip_colors=“${default_skip_colors}”tail=“${default_tail}”show_color_index=“${default_show_color_index}”if [[ ${1} != -* ]]then pod=“${1}”ficontainers=()selector=()regex=‘substring’previous=“${default_previous}”since=“${default_since}”version=“1。6。18-SNAPSHOT”dryrun=falsecluster=“”namespace_arg=“-n ${default_namespace}”usage=“${PROGNAME} [-h] [-c] [-n] [-t] [-l] [-f] [-d] [-P] [-p] [-s] [-b] [-e] [-j] [-k] [-z] [-v] [-r] [-i] —— tail multiple Kubernetes pod logs at the same timewhere: -h, ——help Show this help text。 -c, ——container The name of the container to tail in the pod (if multiple containers are defined in the pod)。 Defaults to all containers in the pod。 Can be used multiple times。 -t, ——context The k8s context。 ex。 int1-context。 Relies on ~/。kube/config for the contexts。 -l, ——selector Label selector。 If used the pod name is ignored。 -n, ——namespace The Kubernetes namespace where the pods are located。 Defaults to \”${default_namespace}\“。 -f, ——follow Specify if the logs should be streamed。 (true|false) Defaults to ${default_follow}。 -d, ——dry-run Print the names of the matched pods and containers, then exit。 -P, ——prefix Specify if add the pod name prefix before each line。 (true|false) Defaults to ${default_prefix}。 -p, ——previous Return logs for the previous instances of the pods, if available。 (true|false) Defaults to ${default_previous}。 -s, ——since Only return logs newer than a relative duration like 5s, 2m, or 3h。 Defaults to ${default_since}。 -b, ——line-buffered This flags indicates to use line-buffered。 (true|false) Defaults to ${default_line_buffered:-false}。 -e, ——regex The type of name matching to use (regex|substring)。 Defaults to ${regex}。 -j, ——jq If your output is json - use this jq-selector to parse it。 Defaults to \”${default_jq_selector}\“。 example: ——jq \”。logger + \\\“ \\\” + 。message\“ -k, ——colored-output Use colored output (pod|line|false)。 pod = only color pod name, line = color entire line, false = don‘t use any colors。 Defaults to ${default_colored_output}。 -z, ——skip-colors Comma-separated list of colors to not use in output。 If you have green foreground on black, this will skip dark grey and some greens: -z 2,8,10 Defaults to: ${default_skip_colors}。 ——timestamps Show timestamps for each log line。 (true|false) Defaults to ${default_timestamps:-false}。 ——tail Lines of recent log file to display。 Defaults to ${default_tail}, showing all log lines。 -v, ——version Prints the kubetail version。 -r, ——cluster The name of the kubeconfig cluster to use。 -i, ——show-color-index Show the color index before the pod name prefix that is shown before each log line。 Normally only the pod name is added as a prefix before each line, for example \”[app-5b7ff6cbcd-bjv8n]\“, but if \”show-color-index\“ is true then color index is added as well: \”[1:app-5b7ff6cbcd-bjv8n]\“。 This is useful if you have color blindness or if you want to know which colors to exclude (see \”——skip-colors\“)。 Defaults to ${default_show_color_index}。examples: ${PROGNAME} my-pod-v1 ${PROGNAME} my-pod-v1 -c my-container ${PROGNAME} my-pod-v1 -t int1-context -c my-container ${PROGNAME} ’(service|consumer|thing)‘ -e regex ${PROGNAME} -l service=my-service ${PROGNAME} ——selector service=my-service ——since 10m ${PROGNAME} ——tail 1”if [ “$#” -ne 0 ]; then while [ “$#” -gt 0 ] do case “$1” in -h|——help) echo “$usage” exit 0 ;; -v|——version) echo “$version” exit 0 ;; -c|——container) containers+=(“$2”) ;; -e|——regex) if [ “$2” = “substring” ]; then regex=“substring” else regex=“regex” fi ;; -t|——context) context=“$2” ;; -r|——cluster) cluster=“——cluster $2” ;; -l|——selector) selector=(——selector “$2”) pod=“” ;; -d|——dry-run) dryrun=true ;; -p|——previous) if [ “$2” = “false” ]; then previous=“false” else previous=“true” fi ;; -s|——since) if [ -z “$2” ]; then since=“${default_since}” else since=“$2” fi ;; -n|——namespace) if [ -z “$2” ]; then # using namespace from context : else namespace_arg=“——namespace $2” fi ;; -f|——follow) if [ “$2” = “false” ]; then follow=“false” else follow=“true” fi ;; -P|——prefix) if [ “$2” = “false” ]; then prefix=“false” else prefix=“true” fi ;; -b|——line-buffered) if [ “$2” = “false” ]; then line_buffered=“” else line_buffered=“| grep - ——line-buffered” fi ;; -k|——colored-output) if [ -z “$2” ]; then colored_output=“${default_colored_output}” else colored_output=“$2” fi ;; -j|——jq) if [ -z “$2” ]; then jq_selector=“${default_jq_selector}” else jq_selector=“$2” fi ;; -z|——skip-colors) if [ -z “$2” ]; then skip_colors=“${default_skip_colors}” else skip_colors=“$2” fi ;; ——timestamps) if [ “$2” = “false” ]; then timestamps=“$1=$2” else timestamps=“$1” fi ;; ——tail) if [ -z “$2” ]; then tail=“${default_tail}” else tail=“$2” fi ;; -i|——show-color-index) if [ -z “$2” ]; then show_color_index=“${default_show_color_index}” else show_color_index=“$2” fi ;; ——) break ;; -*) echo “Invalid option ’$1‘。 Use ——help to see the valid options” >&2 exit 1 ;; # an option argument, continue *) ;; esac shift doneelse echo “$usage” exit 1fi# Join function that supports a multi-character separator (copied from http://stackoverflow。com/a/23673883/398441)function join() { # $1 is sep # $2。。。 are the elements to join local sep=“$1” shift local F=0 for x in “$@” do if [[ F -eq 1 ]] then echo -n “$sep” else F=1 fi echo -n “$x” done echo}# Check if pod query contains a comma and we’ve not specified “regex” explicitly,# if so we convert the pod query string into a regex that matches all pods seperated by the commaif [[ “${pod}” = *“,”* ]] && [ ! “${regex}” == ‘regex’ ]; then # Split the supplied query string (in variable pod) by comma into an array named “pods_to_match” IFS=‘,’ read -r -a pods_to_match <<< “${pod}” # Join all pod names into a string with “。*|。*” as delimiter pod=$(join “。*|。*” “${pods_to_match[@]}”) # Prepend and initial “。*” and and append the last “。*” pod=“。*${pod}。*” # Force the use of regex matching regex=‘regex’figrep_matcher=‘’if [ “${regex}” == ‘regex’ ]; then echo “Using regex ‘${pod}’ to match pods” grep_matcher=‘-E’fi# Get all pods matching the input and put them in an array。 If no input then all pods are matched。matching_pods=(`${KUBECTL_BIN} get pods ${context:+——context=${context}} “${selector[@]}” ${namespace_arg} ${cluster} ——output=jsonpath=‘{。items[*]。metadata。name}’ | xargs -n1 | grep ——color=never $grep_matcher “${pod}”`)matching_pods_size=${#matching_pods[@]}if [ ${matching_pods_size} -eq 0 ]; then echo “No pod exists that matches ${pod}” exit 1ficolor_end=$(tput sgr0)# Wrap all pod names in the “kubectl logs -f=true/false” commanddisplay_names_preview=()pod_logs_commands=()i=0color_index=0function next_col { potential_col=$(($1+1)) [[ $skip_colors =~ (^|,)$potential_col($|,) ]] && echo `next_col $potential_col` || echo $potential_col}# Allows for more colors, this is useful if one tails a lot podsif [ ${colored_output} != “false” ]; then export TERM=xterm-256colorfi# Function that kills all kubectl processes that are started by kubetail in the backgroundfunction kill_kubectl_processes { kill 0}# Invoke the “kill_kubectl_processes” function when the script is stopped (including ctrl+c)# Note that “INT” is not used because if, for example, kubectl cannot find a container# (for example when running “kubetail something -c non_matching”)trap kill_kubectl_processes EXIT# Putting all needed values in a variable so that multiple requests to Kubernetes api can be avoided, thus making it fasterall_pods_containers=$(echo -e `${KUBECTL_BIN} get pods ${namespace_arg} ${context:+——context=${context}} ——output=jsonpath=“{range 。items[*]}{。metadata。name} {。spec[‘containers’, ‘initContainers’][*]。name} \n{end}”`)for pod in ${matching_pods[@]}; do if [ ${#containers[@]} -eq 0 ]; then pod_containers=($(echo -e “$all_pods_containers” | grep $pod | cut -d ‘ ’ -f2- | xargs -n1)) else pod_containers=(“${containers[@]}”) fi for container in ${pod_containers[@]}; do [ ${matching_pods_size} -eq 1 -a ${#pod_containers[@]} -eq 1 ] && single_stream=“true” || single_stream=“false” if [ ${colored_output} == “false” ] || [ ${single_stream} == “true” ]; then color_start=$(tput sgr0) color_index_prefix=“” else color_index=`next_col $color_index` color_start=$(tput setaf $color_index) color_index_prefix=`if [ ${show_color_index} == “true” ]; then echo “${color_index}:”; else echo “”; fi` fi if [ ${#pod_containers[@]} -eq 1 ]; then display_name=“${pod}” else display_name=“${pod} ${container}” fi if [ ${colored_output} == “false” ]; then display_names_preview+=(“${display_name}”) else display_names_preview+=(“$color_index_prefix${color_start}${display_name}${color_end}”) fi if [ ${prefix} == “false” ]; then prefix_line=“” else if [ ${colored_output} == “false” ]; then prefix_line=“[${display_name}] ” else prefix_line=“${color_start}[${color_end}${color_index_prefix}${color_start}${display_name}]${color_end} ” fi fi if [ ${colored_output} == “false” ] || [ ${colored_output} == “pod” ]; then colored_line=“${prefix_line}\$REPLY” else colored_line=“${prefix_line}${color_start}\$REPLY${color_end}” fi kubectl_cmd=“${KUBECTL_BIN} ${context:+——context=${context}} logs ${pod} ${container} -f=${follow} ——previous=${previous} ——since=${since} ——tail=${tail} ${namespace_arg} ${cluster}” colorify_lines_cmd=“while read -r; do echo \”$colored_line\“ | tail -n +1; done” if [ “z” == “z$jq_selector” ]; then logs_commands+=(“${kubectl_cmd} ${timestamps} | ${colorify_lines_cmd}”); else logs_commands+=(“${kubectl_cmd} | jq ——unbuffered -r -R ——stream ‘。 as \$line | try (fromjson | $jq_selector) catch \$line’ | ${colorify_lines_cmd}”); fi # There are only 11 usable colors i=$(( ($i+1)%13 )) donedone# Preview pod colorsecho “Will tail ${#display_names_preview[@]} logs。。。”for preview in “${display_names_preview[@]}”; do echo “$preview”doneif [[ ${dryrun} == true ]];then exit 0fi# Join all log commands into one string separated by “ & ”command_to_tail=$(join “ & ” “${logs_commands[@]}”)# Aggregate all logs and print to stdout# Note that tail +1f doesn‘t work on some Linux distributions so we use this slightly longer alternative# Note that if ——follow=false, then the tail command should also not be followedtail_follow_command=“-f”if [[ ${follow} == false ]];then tail_follow_command=“”fitail ${tail_follow_command} -n +1 <( eval “${command_to_tail}” ) $line_buffered

相關文章

頂部