During test_worker_master_removal, filebeat goes in error and test fails

[Camille Rodriguez]

I ran the tests from https://github.com/charmed-kubernetes/jenkins/blob/master/jobs/integration/validation.py on my deployment and the test test_worker_master_removal failed because filebeat went into error (error on hook "stop"). I did a "juju resolved filebeat/x" and the error went away, the worker unit was able to be removed. This might be a
bug with filebeat, or with the test, I am unsure.

The k8s bundle :
description: A highly-available, production-grade Kubernetes cluster.
series: bionic
variables:
  kubeapi-loadbalancer-vip: &kubeapi-load-balancer-vip "10.1.234.21"
applications:
  easyrsa:
    charm: cs:~containers/easyrsa
    constraints: root-disk=8G zones=MYZONE
    num_units: 1
  etcd:
    charm: cs:etcd
    constraints: cores=4 mem=8G root-disk=100G zones=MYZONE
    num_units: 3
    options:
      # XXX: bug 1826009
      channel: 3.2/stable
  canal:
    charm: cs:~containers/canal
    options:
      cidr: 172.17.248.0/21
    nagios_context: nagios
  canonical-livepatch:
   charm: cs:canonical-livepatch
   options:
     livepatch_key: KEY
  kubeapi-load-balancer:
    charm: cs:~containers/kubeapi-load-balancer
    constraints: cores=2 mem=8G root-disk=50G zones=MYZONE
    expose: true
    num_units: 3
    options:
      ha-cluster-vip: *kubeapi-load-balancer-vip
  hacluster-kubeapi-load-balancer:
    charm: cs:hacluster
    options:
      cluster_count: 3
  kubernetes-master:
    charm: cs:~containers/kubernetes-master
    constraints: cores=4 mem=8G root-disk=100G zones=MYZONE
    num_units: 2
    options:
      allow-privileged: "true"
      channel: 1.17/stable
      service-cidr: 192.152.183.0/24
      # XXX: bug 1841800
      authorization-mode: "RBAC,Node"
  containerd:
    charm: cs:~containers/containerd
    options:
      # XXX: bug 1841438
      # http_proxy: FCE_TEMPLATE
      # https_proxy: FCE_TEMPLATE
  kubernetes-worker:
    charm: cs:~containers/kubernetes-worker
    constraints: cores=8 mem=16G root-disk=100G zones=MYZONE
    expose: true
    num_units: 3
    options:
      allow-privileged: "true"
      channel: 1.17/stable
  filebeat:
    charm: cs:filebeat
    options:
      logpath: '/var/log/*.log'
      kube_logs: True
  telegraf:
    charm: cs:telegraf
  nrpe-host:
    charm: cs:nrpe
    options:
      nagios_hostname_type: "host"
      nagios_host_context: nagios
      xfs_errors: "30"
  landscape-client:
    charm: cs:landscape-client
    options:
      account-name: "standalone"
      origin: ppa:landscape/19.10
      registration-key: key
      disable-unattended-upgrades: true
relations:
- - kubernetes-master:kube-api-endpoint
  - kubeapi-load-balancer:apiserver
- - kubernetes-master:loadbalancer
  - kubeapi-load-balancer:loadbalancer
- - kubeapi-load-balancer
  - hacluster-kubeapi-load-balancer
- - kubernetes-master:kube-control
  - kubernetes-worker:kube-control
- - kubernetes-master:certificates
  - easyrsa:client
- - kubernetes-worker:certificates
  - easyrsa:client
- - kubeapi-load-balancer:certificates
  - easyrsa:client
- - etcd:certificates
  - easyrsa:client
- - kubernetes-master:etcd
  - etcd:db
- - kubernetes-worker:kube-api-endpoint
  - kubeapi-load-balancer:website
- - canal:etcd
  - etcd:db
- - canal:cni
  - kubernetes-master:cni
- - canal:cni
  - kubernetes-worker:cni
- - filebeat:beats-host
  - kubernetes-master:juju-info
- - filebeat:beats-host
  - kubernetes-worker:juju-info
- - filebeat:logstash
  - graylog:beats
- - telegraf:prometheus-client
  - prometheus-target:target
- - kubernetes-master:juju-info
  - telegraf:juju-info
- - kubernetes-worker:juju-info
  - telegraf:juju-info
- - canonical-livepatch
  - etcd
- - canonical-livepatch
  - kubernetes-master
- - canonical-livepatch
  - kubernetes-worker
- - nagios:monitors
  - nrpe-host:monitors
- - easyrsa
  - nrpe-host
- - etcd
  - nrpe-host
- - kubernetes-master
  - nrpe-host
- - containerd
  - kubernetes-master
- - containerd
  - kubernetes-worker
- - landscape-client
  - easyrsa
- - landscape-client
  - etcd
- - landscape-client
  - kubernetes-master
- - landscape-client
  - kubernetes-worker
- - landscape-client
  - kubeapi-load-balancer
- - kubernetes-master:kube-api-endpoint
  - prometheus-target:target
- - kubernetes-master:prometheus
  - prometheus-jobs:manual-jobs
- - kubernetes-master:grafana
  - grafana:dashboards
saas:
  graylog:
    url: foundation_vsphere:admin/lma.graylog-beats
  prometheus-target:
    url: foundation_vsphere:admin/lma.prometheus-target
  prometheus-jobs:
    url: foundation_vsphere:admin/lma.prometheus-jobs
  nagios:
    url: foundation_vsphere:admin/lma.nagios-monitors
  grafana:
    url: foundation_vsphere:admin/lma.grafana-dashboards