Operation
You can run any zuul process with the -f option to make it not daemonize and stay in the foreground, logging to your terminal. It’s a good idea at first to check for issues with your configuration. There’s also a -d option to engage verbose debug logging, but be careful in busy deployments as this can generate very large logs.
To start, simply run:
zuul-scheduler
Before Zuul can run any jobs, it needs to load its configuration, most of which is in the git repositories that Zuul operates on. Start an executor to allow zuul to do that:
zuul-executor
Zuul should now be able to read its configuration from the configured repo and process any jobs defined therein.
Scheduler
Operation
To start the scheduler, run zuul-scheduler
. To stop it, run
zuul-scheduler stop
.
Reconfiguration
Most of Zuul’s configuration is automatically updated as changes to the repositories which contain it are merged. However, Zuul must be explicitly notified of changes to the tenant config file, since it is not read from a git repository. Zuul supports two kinds of reconfigurations.
The full reconfiguration refetches and reloads the configuration of
all tenants. To do so, run zuul-scheduler full-reconfigure
. For
example this can be used to fix eventual configuration inconsistencies
after connection problems with the code hosting system.
To perform the same actions as a full reconfiguration but for a single
tenant, use zuul-scheduler tenant-reconfigure TENANT
(where
TENANT
is the name of the tenant to reconfigure).
The smart reconfiguration reloads only the tenants that changed their
configuration in the tenant config file. To do so, run
zuul-scheduler smart-reconfigure
. In multi tenant systems this can
be much faster than the full reconfiguration so it is recommended to
use the smart reconfiguration after changing the tenant configuration
file.
The tenant-reconfigure
and smart-reconfigure
commands should
only be run on a single scheduler. Other schedulers will see any
changes to the configuration stored in ZooKeeper and automatically
update their configuration in the background without interrupting
processing.
Backup and Restoration
While all of Zuul’s component services are designed to be run in a resilient active-active clustered deployment, a good disaster recovery plan should include backing up critical data. At a minimum, the randomly-generated project keys used for encryption of job secrets and SSH access should be backed up, as they cannot be recreated if lost. Zuul stores these keys in a keystore in ZooKeeper which is inconvenient to back up directly, but provides an administrative tool to export these keys to and import them from a local directory.
It’s highly recommended to set up periodic automation for dumping such an export to a secure location (for example, on the filesystem of each Zuul Scheduler) for use in a disaster where all ZooKeeper content is lost. You may also consider configuring a safe remote backup of these files with the tool of your choice, but be aware that they are potentially sensitive since anyone who gains access to them could decrypt job secrets or access protected systems which have been instructed to trust those keys.
Note that the exported keys are symmetrically encrypted with the same keystore.password which is used for encrypting and decrypting the copy of them in ZooKeeper, because its the encrypted versions of the keys which are exported and imported. Someone with access to the keys would also need a copy of the keystore.password from Zuul’s configuration, so for security-sensitive environments you may not want to back them up together. Conversely, if you lose the keystore.password then you also lose the use of the project keys in the keystore and any exports, so you will likely want to make sure you keep a secured copy of it somewhere as well in the event your server configuration is lost.
Merger
Operation
To start the merger, run zuul-merger
.
In order to stop the merger and under normal circumstances it is
best to pause and wait for all currently running tasks to finish
before stopping it. To do so run zuul-merger pause
.
To stop the merger, run zuul-merger stop
. This will wait for any
currently running merge task to complete before exiting. As a result
this is always a graceful way to stop the merger.
zuul-merger graceful
is an alias for zuul-merger stop
to make
this consistent with the executor.
Executor
Operation
To start the executor, run zuul-executor
.
There are several commands which can be run to control the executor’s behavior once it is running.
To pause the executor and prevent it from running new jobs you can
run zuul-executor pause
.
To cause the executor to stop accepting new jobs and exit when all running
jobs have finished you can run zuul-executor graceful
. Under most
circumstances this will be the best way to stop Zuul.
To stop the executor immediately, run zuul-executor stop
. Jobs that were
running on the stopped executor will be rescheduled on other executors.
The executor normally responds to a SIGTERM
signal in the same way
as the graceful
command, however you can change this behavior to match
stop
with the executor.sigterm_method setting.
To enable or disable running Ansible in verbose mode (with the
-vvv
argument to ansible-playbook) run zuul-executor verbose
and zuul-executor unverbose
.
Ansible and Python 3
As noted above, the executor runs Ansible playbooks against the remote
node(s) allocated for the job. Since part of executing playbooks on
remote hosts is running Python scripts on them, Ansible needs to know
what Python interpreter to use on the remote host. With older
distributions, /usr/bin/python2
was a generally sensible choice.
However, over time a heterogeneous Python ecosystem has evolved where
older distributions may only provide Python 2, most provide a mixed
2/3 environment and newer distributions may only provide Python 3 (and
then others like RHEL8 may even have separate “system” Python versions
to add to confusion!).
Ansible’s ansible_python_interpreter
variable configures the path
to the remote Python interpreter to use during playbook execution.
This value is set by Zuul from the python-path
specified for the
node by Nodepool; see the nodepool configuration documentation.
This defaults to auto
, where Ansible will automatically discover
the interpreter available on the remote host. However, this setting
only became available in Ansible >=2.8, so Zuul will translate
auto
into the old default of /usr/bin/python2
when configured
to use older Ansible versions.
Thus for modern Python 3-only hosts no further configuration is needed
when using Ansible >=2.8 (e.g. Fedora, Bionic onwards). If using
earlier Ansible versions you may need to explicitly set the
python-path
if /usr/bin/python2
is not available on the node.
Ansible roles/modules which include Python code are generally Python 3 safe now, but there is still a small possibility of incompatibility. See also the Ansible Python 3 support page.
Log streaming
The log streaming service enables Zuul to show the live status of
long-running shell
or command
tasks. The server side is setup
by the zuul_console:
task built-in to Zuul’s Ansible installation.
The executor requires the ability to communicate with this server on
the job nodes via port 19885
for this to work.
The log streaming service spools command output via files on the job
node in the format /tmp/console-<uuid>-<task_id>-<host>.log
. By
default, it will clean these files up automatically.
Occasionally, a streaming file may be left if a job is interrupted. These may be safely removed after a short period of inactivity with a command such as
find /tmp -maxdepth 1 -name 'console-*-*-<host>.log' -mtime +2 -delete
If the executor is unable to reach port 19885
(for example due to
firewall rules), or the zuul_console
daemon can not be run for
some other reason, the command to clean these spool files will not be
processed and they may be left behind; on an ephemeral node this is
not usually a problem, but on a static node these files will persist.
In this situation, Zuul can be instructed to not to create any spool
files for shell
and command
tasks via setting
zuul_console_disabled: True
(usually via a global host variable in
inventory). Live streaming of shell
and command
calls will of
course be unavailable in this case, but no spool files will be
created.
For Kubernetes-based job nodes the connection from the executor to the
zuul_console
daemon is established by using kubectl port-forward
to forward a local port to the appropriate port on the pod containing
the job node. If the Kubernetes user is not bound to a role that has
authorization for port-forwarding, this will prevent connection to
the zuul_console
daemon.
Web Server
Operation
To start the web server, run zuul-web
. To stop it, kill the
PID which was saved in the pidfile specified in the configuration.
Finger Gateway
Operation
To start the finger gateway, run zuul-fingergw
. To stop it, kill the
PID which was saved in the pidfile specified in the configuration.