Issue

• What services should run on my CJE1 workers and controllers?

Environment

• CloudBees Jenkins Enterprise

Resolution

Important note

The following list is here to give an overview of the services that are supposed to run.
In no way you should manually start services on your controllers and workers.
In case of doubt, you should try to run a dna init <worker|controller>, that will restart the necessary services.

Services on a controller

The output of service --status-all should look like:

service --status-all
 [ + ]  acpid
 [ - ]  apparmor
 [ ? ]  apport
 [ + ]  atd
 [ + ]  chrony
 [ ? ]  console-setup
 [ + ]  cron
 [ ? ]  cryptdisks
 [ ? ]  cryptdisks-early
 [ - ]  dbus
 [ ? ]  dns-clean
 [ + ]  docker
 [ + ]  friendly-recovery
 [ - ]  grub-common
 [ ? ]  irqbalance
 [ ? ]  killprocs
 [ ? ]  kmod
 [ - ]  landscape-client
 [ + ]  marathon
 [ ? ]  networking
 [ ? ]  ondemand
 [ ? ]  open-vm-tools
 [ ? ]  pppd-dns
 [ - ]  procps
 [ ? ]  rc.local
 [ + ]  resolvconf
 [ + ]  rpcbind
 [ - ]  rsync
 [ + ]  rsyslog
 [ ? ]  screen-cleanup
 [ ? ]  sendsigs
 [ + ]  ssh
 [ - ]  sudo
 [ - ]  topbeat
 [ + ]  udev
 [ ? ]  umountfs
 [ ? ]  umountnfs.sh
 [ ? ]  umountroot
 [ - ]  unattended-upgrades
 [ - ]  urandom
 [ - ]  x11-common
 [ + ]  zookeeper

In addition, you want to review the status of the mesos-master and mesos-slave services, by running initctl list |grep mesos:

mesos-master start/running, process XXXXX
mesos-slave stop/waiting

It is crucial that only the master runs on a controller.

Services on a worker

The output of service --status-all should look like:

 [ + ]  acpid
 [ - ]  apparmor
 [ ? ]  apport
 [ + ]  atd
 [ + ]  chrony
 [ ? ]  console-setup
 [ + ]  cron
 [ ? ]  cryptdisks
 [ ? ]  cryptdisks-early
 [ - ]  dbus
 [ ? ]  dns-clean
 [ + ]  docker
 [ + ]  friendly-recovery
 [ - ]  grub-common
 [ ? ]  irqbalance
 [ ? ]  killprocs
 [ ? ]  kmod
 [ - ]  landscape-client
 [ - ]  marathon
 [ ? ]  networking
 [ - ]  ntp
 [ ? ]  ondemand
 [ ? ]  open-vm-tools
 [ ? ]  pppd-dns
 [ - ]  procps
 [ ? ]  rc.local
 [ + ]  resolvconf
 [ + ]  rpcbind
 [ - ]  rsync
 [ + ]  rsyslog
 [ ? ]  screen-cleanup
 [ ? ]  sendsigs
 [ + ]  ssh
 [ - ]  sudo
 [ - ]  topbeat
 [ + ]  udev
 [ ? ]  umountfs
 [ ? ]  umountnfs.sh
 [ ? ]  umountroot
 [ - ]  unattended-upgrades
 [ - ]  urandom
 [ - ]  x11-common
 [ - ]  zookeeper

In addition, you want to review the status of the mesos-master and mesos-slave services, by running initctl list |grep mesos:

mesos-master stop/waiting
mesos-slave start/running, process XXXXX

It is crucial that only the agent runs on a worker.

Tested product/plugin versions

• CloudBees Jenkins Enterprise 1.11.32