6 kutaannada nidaamka madadaalada ee hawlgalka Kubernetes [iyo xalkooda]

Sanadihii la soo dhaafay ee isticmaalka Kubernetes ee wax soo saarka, waxaan uruurinay sheekooyin badan oo xiiso leh oo ku saabsan sida cayayaanka qaybaha nidaamka kala duwan ay u horseedeen cawaaqib xumo iyo / ama aan la fahmi karin oo saameeya hawlgalka weelasha iyo boodhka. Maqaalkan waxaan ku soo xulnay qaar ka mid ah kuwa ugu caansan ama xiisaha leh. Xitaa haddii aadan weligaa nasiib u yeelan inaad la kulanto xaaladahan oo kale, akhrinta ku saabsan sheekooyinka baarayaasha gaagaaban - gaar ahaan "gacanta koowaad" - had iyo jeer waa mid xiiso leh, miyaanay ahayn?

Sheekada 1. Supercronic iyo Docker oo laadlaadsan

Mid ka mid ah kooxaha, waxaan si xilliyo ah ku helnay Docker barafaysan, kaas oo farageliyay shaqada caadiga ah ee kooxda. Isla mar ahaantaana, kuwan soo socda ayaa lagu arkay diiwaanka Docker:

level=error msg="containerd: start init process" error="exit status 2: "runtime/cgo: pthread_create failed: No space left on device
SIGABRT: abort
PC=0x7f31b811a428 m=0

goroutine 0 [idle]:

goroutine 1 [running]:
runtime.systemstack_switch() /usr/local/go/src/runtime/asm_amd64.s:252 fp=0xc420026768 sp=0xc420026760
runtime.main() /usr/local/go/src/runtime/proc.go:127 +0x6c fp=0xc4200267c0 sp=0xc420026768
runtime.goexit() /usr/local/go/src/runtime/asm_amd64.s:2086 +0x1 fp=0xc4200267c8 sp=0xc4200267c0

goroutine 17 [syscall, locked to thread]:
runtime.goexit() /usr/local/go/src/runtime/asm_amd64.s:2086 +0x1

…

Waxa ugu xiisaha badan qaladkan waa fariinta: pthread_create failed: No space left on device. Waxbarasho Degdeg ah dukumentiyo sharaxay in Docker aanu fargeeto habsocodka, taas oo ah sababta ay xilliyada qaarkood u barafoobaan.

Kormeerka, sawirka soo socdaa wuxuu u dhigmaa waxa dhacaya:

Xaalad la mid ah ayaa lagu arkay qanjidhada kale:

Isla qanjidhada ayaanu ku aragnaa:

root@kube-node-1 ~ # ps auxfww | grep curl -c
19782
root@kube-node-1 ~ # ps auxfww | grep curl | head
root     16688  0.0  0.0      0     0 ?        Z    Feb06   0:00      |       _ [curl] <defunct>
root     17398  0.0  0.0      0     0 ?        Z    Feb06   0:00      |       _ [curl] <defunct>
root     16852  0.0  0.0      0     0 ?        Z    Feb06   0:00      |       _ [curl] <defunct>
root      9473  0.0  0.0      0     0 ?        Z    Feb06   0:00      |       _ [curl] <defunct>
root      4664  0.0  0.0      0     0 ?        Z    Feb06   0:00      |       _ [curl] <defunct>
root     30571  0.0  0.0      0     0 ?        Z    Feb06   0:00      |       _ [curl] <defunct>
root     24113  0.0  0.0      0     0 ?        Z    Feb06   0:00      |       _ [curl] <defunct>
root     16475  0.0  0.0      0     0 ?        Z    Feb06   0:00      |       _ [curl] <defunct>
root      7176  0.0  0.0      0     0 ?        Z    Feb06   0:00      |       _ [curl] <defunct>
root      1090  0.0  0.0      0     0 ?        Z    Feb06   0:00      |       _ [curl] <defunct>

Waxaa soo baxday in hab-dhaqankani uu yahay natiijada ka soo baxda boodhka oo ku shaqeeya supercronic (A Go utility oo aan u isticmaalno si aan ugu socodsiino shaqooyinka cron ee pods):

 _ docker-containerd-shim 833b60bb9ff4c669bb413b898a5fd142a57a21695e5dc42684235df907825567 /var/run/docker/libcontainerd/833b60bb9ff4c669bb413b898a5fd142a57a21695e5dc42684235df907825567 docker-runc
|   _ /usr/local/bin/supercronic -json /crontabs/cron
|       _ /usr/bin/newrelic-daemon --agent --pidfile /var/run/newrelic-daemon.pid --logfile /dev/stderr --port /run/newrelic.sock --tls --define utilization.detect_aws=true --define utilization.detect_azure=true --define utilization.detect_gcp=true --define utilization.detect_pcf=true --define utilization.detect_docker=true
|       |   _ /usr/bin/newrelic-daemon --agent --pidfile /var/run/newrelic-daemon.pid --logfile /dev/stderr --port /run/newrelic.sock --tls --define utilization.detect_aws=true --define utilization.detect_azure=true --define utilization.detect_gcp=true --define utilization.detect_pcf=true --define utilization.detect_docker=true -no-pidfile
|       _ [newrelic-daemon] <defunct>
|       _ [curl] <defunct>
|       _ [curl] <defunct>
|       _ [curl] <defunct>
…

Dhibaatadu waa tan: marka hawsha lagu wado si heer sare ah, geeddi-socodku waa uu dhalay si sax ah uma joojin karo, isu rogaya zombie.

tacliiq: Si aad u noqoto mid sax ah, hababka waxaa dhaley hawlaha cron, laakiin supercronic ma aha nidaamka init mana "qaadan karo" hababka ay carruurtu dhaleen. Marka calaamadaha SIGHUP ama SIGTERM la kiciyo, looma gudbiyo habsocodka ubadka, taasoo keentay in hab-socodka ilmuhu aanu joojin oo uu ku sii jiro heerka zombie. Waxaad ka akhrisan kartaa wax badan oo ku saabsan waxaas oo dhan, tusaale ahaan, gudaha maqaal noocaas ah.

Waxaa jira dhowr siyaabood oo lagu xalliyo dhibaatooyinka:

1. Ka-hortagga ku-meel-gaarka ah - kordhi tirada PID-yada ee nidaamka hal mar oo keliya:

          /proc/sys/kernel/pid_max (since Linux 2.5.34)
                 This file specifies the value at which PIDs wrap around (i.e., the value in this file is one greater than the maximum PID).  PIDs greater than this  value  are  not  allo‐
                 cated;  thus, the value in this file also acts as a system-wide limit on the total number of processes and threads.  The default value for this file, 32768, results in the
                 same range of PIDs as on earlier kernels

2. Ama u billow hawlaha supercronic si toos ah maaha, laakiin adoo isticmaalaya isku mid tini, kaas oo awood u leh inuu si sax ah u joojiyo hababka oo aan dhalin zombies.

Sheekada 2. "Zombies" marka la tirtirayo koox koox

Kubelet wuxuu bilaabay cunista CPU badan:

Qofna sidan ma jecla, markaa waannu hubaynay kaamil ah waxayna bilaabeen inay wax ka qabtaan dhibaatada. Natiijadii baadhista ayaa noqotay sidan.

• Kubelet waxay ku bixisaa in ka badan seddex meelood meel waqtigeeda CPU ka soo jiidashada xogta xusuusta dhammaan kooxaha:

  

• Liiska boostada ee horumarinta kernel-ka waxaad ka heli kartaa ka hadalka dhibaatada. Marka la soo koobo, dulucdu waxay ku soo degtaa sidan: faylalka kala duwan ee tmpfs iyo waxyaabo kale oo la mid ah ayaan gabi ahaanba laga saarin nidaamka marka la tirtirayo koox-koox, waxa loogu yeero memcg Zombie. Mar dhow ama hadhowba waa laga tirtiri doonaa khasnadda bogga, laakiin waxaa jira kayd badan oo server-ka ah oo kernel-ku ma arko barta wakhtiga lagu luminayo tirtirka. Taasi waa sababta ay u sii ururinayaan. Maxay tani xitaa u dhacaysaa? Kani waa adeege leh shaqooyin cron ah oo si joogto ah u abuura shaqooyin cusub, oo ay la socdaan qaybo cusub. Markaa, kooxo cusub ayaa loo abuuray weelasha ku jira, kuwaas oo isla markiiba la tirtiro.
• Waa maxay sababta lataliyaha kubelet uu waqti badan u lumiyo? Tani way fududahay in lagu arko fulinta ugu fudud time cat /sys/fs/cgroup/memory/memory.stat. Haddii mashiinka caafimaadka qaba qalliinku uu qaato 0,01 ilbiriqsi, ka dibna cron02 dhibaatada leh waxay qaadataa 1,2 ilbiriqsi. Shaydu waa cAdvisor, oo akhriya xogta sysfs si tartiib tartiib ah, wuxuu isku dayaa inuu tixgeliyo xusuusta loo isticmaalo kooxaha zombies.
• Si xoog looga saaro zombies-ka, waxaan isku daynay inaan nadiifinno khasnadaha sida lagu taliyay LKML: sync; echo 3 > /proc/sys/vm/drop_caches, - laakiin kernel ayaa noqday mid aad u dhib badan oo shil baabuur ku dhacay.

Maxaa la sameeyaa? Dhibaatadu waa la hagaajinayaa (galan, oo sharaxaad ka eeg fariinta sii daynta) Cusboonaysiinta kernel-ka Linux ilaa version 4.16.

Taariikhda 3. Nidaamka iyo buurtiisa

Mar labaad, kubelet-ku wuxuu cunaa kheyraad aad u badan qanjidhada qaar, laakiin markan waxay cunaysaa xasuus aad u badan:

Waxaa soo baxday in ay jirto dhibaato ku saabsan nidaamka loo isticmaalo nidaamka systemd-ka Ubuntu 16.04, waxayna dhacdaa marka la maamulayo qalabka loo sameeyay isku xirka subPath Laga soo bilaabo ConfigMap's ama sirta. Ka dib markii boodhka uu dhammeeyo shaqadiisa Adeegga systemd-ka ah iyo buurta adeegeedu way hadhaysaa nidaamka. Muddo ka dib, tiro aad u badan ayaa urura. Waxaa jira xitaa arrimo mawduucan ku saabsan:

1. #5916;
2. kubernetes #57345.

kan ugu dambeeya kaas oo tilmaamaya PR in systemd: #7811 (arrinta ku jirta systemd - #7798).

Dhibaatadu ma jirto hadda Ubuntu 18.04, laakiin haddii aad rabto inaad sii wadato isticmaalka Ubuntu 16.04, waxaa laga yaabaa inaad u aragto xalkayaga ku saabsan mowduucan mid waxtar leh.

Markaa waxaanu samaynay DaemonSet-ka soo socda:

---
apiVersion: extensions/v1beta1
kind: DaemonSet
metadata:
  labels:
    app: systemd-slices-cleaner
  name: systemd-slices-cleaner
  namespace: kube-system
spec:
  updateStrategy:
    type: RollingUpdate
  selector:
    matchLabels:
      app: systemd-slices-cleaner
  template:
    metadata:
      labels:
        app: systemd-slices-cleaner
    spec:
      containers:
      - command:
        - /usr/local/bin/supercronic
        - -json
        - /app/crontab
        Image: private-registry.org/systemd-slices-cleaner/systemd-slices-cleaner:v0.1.0
        imagePullPolicy: Always
        name: systemd-slices-cleaner
        resources: {}
        securityContext:
          privileged: true
        volumeMounts:
        - name: systemd
          mountPath: /run/systemd/private
        - name: docker
          mountPath: /run/docker.sock
        - name: systemd-etc
          mountPath: /etc/systemd
        - name: systemd-run
          mountPath: /run/systemd/system/
        - name: lsb-release
          mountPath: /etc/lsb-release-host
      imagePullSecrets:
      - name: antiopa-registry
      priorityClassName: cluster-low
      tolerations:
      - operator: Exists
      volumes:
      - name: systemd
        hostPath:
          path: /run/systemd/private
      - name: docker
        hostPath:
          path: /run/docker.sock
      - name: systemd-etc
        hostPath:
          path: /etc/systemd
      - name: systemd-run
        hostPath:
          path: /run/systemd/system/
      - name: lsb-release
        hostPath:
          path: /etc/lsb-release

... waxayna isticmaashaa qoraalkan soo socda:

#!/bin/bash

# we will work only on xenial
hostrelease="/etc/lsb-release-host"
test -f ${hostrelease} && grep xenial ${hostrelease} > /dev/null || exit 0

# sleeping max 30 minutes to dispense load on kube-nodes
sleep $((RANDOM % 1800))

stoppedCount=0
# counting actual subpath units in systemd
countBefore=$(systemctl list-units | grep subpath | grep "run-" | wc -l)
# let's go check each unit
for unit in $(systemctl list-units | grep subpath | grep "run-" | awk '{print $1}'); do
  # finding description file for unit (to find out docker container, who born this unit)
  DropFile=$(systemctl status ${unit} | grep Drop | awk -F': ' '{print $2}')
  # reading uuid for docker container from description file
  DockerContainerId=$(cat ${DropFile}/50-Description.conf | awk '{print $5}' | cut -d/ -f6)
  # checking container status (running or not)
  checkFlag=$(docker ps | grep -c ${DockerContainerId})
  # if container not running, we will stop unit
  if [[ ${checkFlag} -eq 0 ]]; then
    echo "Stopping unit ${unit}"
    # stoping unit in action
    systemctl stop $unit
    # just counter for logs
    ((stoppedCount++))
    # logging current progress
    echo "Stopped ${stoppedCount} systemd units out of ${countBefore}"
  fi
done

... waxayna ku socotaa 5-tii daqiqo kasta iyadoo la isticmaalayo supercronic hore loo sheegay. Dockerfilekeedu wuxuu u eg yahay sidan:

FROM ubuntu:16.04
COPY rootfs /
WORKDIR /app
RUN apt-get update && 
    apt-get upgrade -y && 
    apt-get install -y gnupg curl apt-transport-https software-properties-common wget
RUN add-apt-repository "deb [arch=amd64] https://download.docker.com/linux/ubuntu xenial stable" && 
    curl -fsSL https://download.docker.com/linux/ubuntu/gpg | apt-key add - && 
    apt-get update && 
    apt-get install -y docker-ce=17.03.0*
RUN wget https://github.com/aptible/supercronic/releases/download/v0.1.6/supercronic-linux-amd64 -O 
    /usr/local/bin/supercronic && chmod +x /usr/local/bin/supercronic
ENTRYPOINT ["/bin/bash", "-c", "/usr/local/bin/supercronic -json /app/crontab"]

Sheekada 4-aad

Waxaa la ogaaday in: haddii aan ku dhejineyno noodhka oo sawirkiisa la soo saaro wakhti aad u dheer, ka dibna qayb kale oo "ku dhufatey" isla noodhka ayaa si fudud u noqon doonta. ma bilaabo inuu jiido sawirka boodhka cusub. Taa beddelkeeda, waxay sugaysaa ilaa sawirka boodhka hore la jiido. Natiijadu waxay tahay, boodh hore loo qorsheeyay oo sawirkeeda la soo dejin karayay hal daqiiqo gudaheed waxay ku dambayn doontaa heerka containerCreating.

Dhacdooyinka ayaa u ekaan doona sidatan:

Normal  Pulling    8m    kubelet, ip-10-241-44-128.ap-northeast-1.compute.internal  pulling image "registry.example.com/infra/openvpn/openvpn:master"

Waxay soo baxaysaa in hal sawir oo ka yimid diiwaanka gaabis ah ayaa xannibi kara keenista meel kasta.

Nasiib darro, ma jiraan siyaabo badan oo looga baxo xaaladda:

1. Isku day inaad si toos ah ugu isticmaasho diiwaanka Docker gaaga kutlada ama si toos ah ula socoto kooxda (tusaale, GitLab Registry, Nexus, iwm.);
2. Isticmaal adeegyada sida kraken.

Sheekada 5. Nodes waxay laalaadaan xusuusta la'aanta awgeed

Inta lagu guda jiro hawlgalka codsiyada kala duwan, waxaan sidoo kale la kulannay xaalad halkaas oo qanjidhada gabi ahaanba joojiso in la heli karo: SSH kama jawaabto, dhammaan daemons la socodka dhaco off, ka dibna ma jiraan wax (ama ku dhawaad ​​waxba) anomalous ee lo'da.

Waxaan kuugu sheegi doonaa sawirada adoo isticmaalaya tusaale hal noode ah oo MongoDB ay ka shaqeyso.

Tani waa waxa dusha sare u eg si ay u shilalka:

Oo sida tan - после shilalka:

Korjoogteynta, waxaa sidoo kale jira bood fiiqan, kaas oo noodu uu joojinayo in la helo:

Haddaba, shaashadaha laga soo qaaday waxaa iska cad in:

1. RAM ee mishiinka ayaa ku dhow dhamaadka;
2. Waxaa jira bood fiiqan isticmaalka RAM, ka dib markii gelitaanka mashiinka oo dhan uu si lama filaan ah u naafo yahay;
3. Hawl weyn ayaa ku timaada Mongo, taas oo ku qasbeysa habka DBMS inuu isticmaalo xasuus badan oo si firfircoon uga akhriyo diskka.

Waxaa soo baxday in haddii gudaha Linux Xusuusta xorta ah way dhammaataa (cadaadis xusuusta ah ayaa dhacda) mana jiro wax is-beddel ah, ka dibna si ay u Marka uu dilaaga OOM yimaado, waxa dhici karta in fal dheeli tiran uu soo baxo inta u dhaxaysa ku tuurista boggaga khasnadda bogga oo dib loogu soo celiyo saxanka. Tan waxa sameeya kswapd, kaas oo si geesinimo leh u xoreeya inta badan bogag xusuusta ah ee suurtogalka ah qaybinta dambe.

Nasiib darro, oo leh culays I/O weyn oo ay weheliso qadar yar oo xusuusta bilaashka ah, kswapd waxay noqotaa cirifka nidaamka oo dhan, waayo way ku xidhan yihiin oo dhan qoondaynta (ciladaha bogga) ee bogagga xusuusta ee nidaamka. Tani waxay socon kartaa muddo aad u dheer haddii hab-raacyadu aysan rabin inay isticmaalaan xusuusta, laakiin ay ku go'an yihiin cidhifyada OOM-dilaha yaamayska.

Su'aasha dabiiciga ah waxay tahay: muxuu dilaa OOM u soo daahaa? Dib-u-eegisteeda hadda, dilaaga OOM waa doqonimo aad u xun: waxay dili doontaa habka kaliya marka isku dayga qoondaynta bogga xusuusta uu guuldareysto, i.e. haddii cilada boggu fashilanto. Tani ma dhacayso wakhti dheer, sababtoo ah kswapd waxay si geesinimo leh u xoraysaa boggaga xusuusta, iyada oo la daadinayo khasnadda bogga (dhammaan diskka I/O ee nidaamka, dhab ahaantii) dib ugu celinta diskka. Faahfaahin dheeraad ah, oo leh sharaxaad tillaabooyinka loo baahan yahay si loo baabi'iyo dhibaatooyinka noocaas ah ee kernel, waad akhrin kartaa halkan.

Dhaqankan waa in la hagaajiyaa oo leh xudunta Linux 4.6 +.

Sheekada 6-aad

Kutlooyinka qaarkood, oo ay ka jiraan dhabanno badan oo shaqeynaya, waxaan bilownay inaan ogaano in intooda badan ay "ku dheggan yihiin" muddo aad u dheer gobolka. Pending, in kasta oo weelasha Docker-ka laftoodu ay horeba ugu shaqaynayaan qanjidhada oo lagu shaqayn karo gacanta.

Intaa waxaa dheer, in describe wax qaldan ma jiraan:

  Type    Reason                  Age                From                     Message
  ----    ------                  ----               ----                     -------
  Normal  Scheduled               1m                 default-scheduler        Successfully assigned sphinx-0 to ss-dev-kub07
  Normal  SuccessfulAttachVolume  1m                 attachdetach-controller  AttachVolume.Attach succeeded for volume "pvc-6aaad34f-ad10-11e8-a44c-52540035a73b"
  Normal  SuccessfulMountVolume   1m                 kubelet, ss-dev-kub07    MountVolume.SetUp succeeded for volume "sphinx-config"
  Normal  SuccessfulMountVolume   1m                 kubelet, ss-dev-kub07    MountVolume.SetUp succeeded for volume "default-token-fzcsf"
  Normal  SuccessfulMountVolume   49s (x2 over 51s)  kubelet, ss-dev-kub07    MountVolume.SetUp succeeded for volume "pvc-6aaad34f-ad10-11e8-a44c-52540035a73b"
  Normal  Pulled                  43s                kubelet, ss-dev-kub07    Container image "registry.example.com/infra/sphinx-exporter/sphinx-indexer:v1" already present on machine
  Normal  Created                 43s                kubelet, ss-dev-kub07    Created container
  Normal  Started                 43s                kubelet, ss-dev-kub07    Started container
  Normal  Pulled                  43s                kubelet, ss-dev-kub07    Container image "registry.example.com/infra/sphinx/sphinx:v1" already present on machine
  Normal  Created                 42s                kubelet, ss-dev-kub07    Created container
  Normal  Started                 42s                kubelet, ss-dev-kub07    Started container

Qaar ka mid ah qodista ka dib, waxaan samaynay malo ah in kubelet si fudud ma haysto wakhti ay ku soo dirto dhammaan macluumaadka ku saabsan xaaladda galmoodka iyo imtixaanada noolaanshaha/ diyaargarowga server-ka API.

Ka dib markaan baranay caawinta, waxaan helnay cabbirada soo socda:

--kube-api-qps - QPS to use while talking with kubernetes apiserver (default 5)
--kube-api-burst  - Burst to use while talking with kubernetes apiserver (default 10) 
--event-qps - If > 0, limit event creations per second to this value. If 0, unlimited. (default 5)
--event-burst - Maximum size of a bursty event records, temporarily allows event records to burst to this number, while still not exceeding event-qps. Only used if --event-qps > 0 (default 10) 
--registry-qps - If > 0, limit registry pull QPS to this value.
--registry-burst - Maximum size of bursty pulls, temporarily allows pulls to burst to this number, while still not exceeding registry-qps. Only used if --registry-qps > 0 (default 10)

Sida muuqata, qiyamka caadiga ah aad bay u yar yihiin, iyo 90% waxay daboolaan dhammaan baahiyaha ... Si kastaba ha ahaatee, xaaladdeena tani kuma filna. Sidaa darteed, waxaanu dejinay qiyamka soo socda:

--event-qps=30 --event-burst=40 --kube-api-burst=40 --kube-api-qps=30 --registry-qps=30 --registry-burst=40

... oo dib u bilaabay kubelets, ka dib waxaan ku aragnay sawirka soo socda ee garaafyada wicitaanada server-ka API:

... iyo haa, wax walba waxay bilaabeen inay duulaan!

PS

Caawintooda aruurinta cayayaanka iyo diyaarinta maqaalkan, waxaan u mahadcelinayaa injineerada badan ee shirkadeena, gaar ahaan saaxiibkay kooxda R&D Andrey Klimentyev (zuzzas).

PPS

Sidoo kale ka akhri boggayaga:

• «kubectl-debug plugin si loogu saxo Kubernetes pods".
• Kubernetes talooyinka & tabaha loop:
  • «Ku wareejinta agabka ku jira kutlada kuna wareejinaya maamulka Helm 2";
  • «Ku saabsan qoondaynta noodhka iyo culayska ku jira codsiga shabakadda";
  • «Gelitaanka goobaha dev";
  • «Dedejinta bootstrap-ka kaydka xogta waaweyn".

Source: www.habr.com