From Wikitech
Jump to navigation Jump to search


Kubernetes (often abbreviated k8s) is a platform for running containers. It is used in Toolforge to isolate Tools from each other and allow distributing Tools across a pool of servers.

Kubernetes webservices

The Toolforge webservice command has a --backend=kubernetes mode that will start, stop, and restart containers designed to run web services for various languages. See our Webservice help for more details.

Kubernetes backend has the following options:

  -m MEMORY, --mem MEMORY
                        Set higher Kubernetes memory limit
  -c CPU, --cpu CPU     Set a higher Kubernetes cpu limit
  -r REPLICAS, --replicas REPLICAS
                        Set the number of pod replicas to use

Kubernetes cronjobs

It is possible to run cron jobs on Kubernetes (see upstream documentation for a full description).

Example cronjob.yaml

Wikiloveslove is a Python 3.7 bot that runs in a Kubernetes deployment. The cronjobs.yaml file that it uses to tell Kubernetes how to start and schedule the bot is reproduced below.

Create the CronJob object in your tool's Kubernetes namespace using kubectl:

$ kubectl apply --validate=true -f $HOME/cronjobs.yaml
cronjob.batch/CRONJOB-NAME configured

After creating the cronjob you can create a test job with kubectl create job --from=cronjob/CRONJOB-NAME test to immediately trigger the cronjob and then access the logs as usual with kubectl logs job/test -f to debug.

Kubernetes continuous jobs

The basic unit of managing execution on a Kubernetes cluster is called a "deployment". Each deployment is described with a YAML configuration file which describes the container images to be started ("pods" in the Kubernetes terminology) and commands to be run inside them after the container is initialized. A deployment also specifies where the pods run and what external resources are connected to them. The upstream documentation is comprehensive.

Example deployment.yaml

Stashbot is a Python 3.7 irc bot that runs in a Kubernetes deployment. The deployment.yaml file that it uses to tell Kubernetes how to start the bot is reproduced below. This deployment is launched using a wrapper script which runs kubectl create --validate=true -f /data/project/stashbot/etc/deployment.yaml.

This deployment:

  • Uses the 'tool-stashbot' namespace that the tool is authorized to control
  • Creates a container using the 'latest' version of the '' Docker image.
  • Runs the command /data/project/stashbot/bin/ run when the container starts.
  • Mounts the /data/project/stashbot/ NFS directory as /data/project/stashbot/ inside the container.

Monitoring your jobs

You can see which jobs you have running with kubectl get pods. Using the name of the pod, you can see the logs with kubectl logs <pod-name>.

To restart a failing pod, use kubectl delete <pod-name>. If you need to kill it entirely, find the deployment name with kubectl get deployment, and delete it with kubectl delete deployment <deployment-name>.


Each tool has been granted control of a Kubernetes "namespace". Your tool can only create and control objects in its namespace. A tool's namespace is the same as the tool's name with "tool-" appended to the beginning (e.g. tool-admin, tool-stashbot, tool-hay, etc).

Quotas and Resources

On the Kubernetes cluster, all containers run with CPU and RAM limits set, just like jobs on the Gridengine cluster. Defaults are set at 0.5 CPU and 512Mi of memory per container. Users can adjust these up to the highest level allowed without any help from an administrator (the top limit is set at 1 CPU and 4GiB of memory) with command line arguments to the webservice command (--cpu and --mem) or properly formatted Kubernetes YAML specifications for your pod's resources fields for advanced users.

The Toolforge admin team encourages you to try running your webservice with the defaults before deciding that you need more resources. We believe that most PHP and Python3 webservices will work as expected with the lower values. Java webservices will almost certainly need higher limits due to the nature of running a JVM.

Namespace-wide quotas

Your entire tool account can only consume so many cluster resources. The cluster places quota limits on an entire namespace which determine how many pods can be used, how many service ports can be exposed, total memory, total CPU, and others. The default limits for a tool's entire namespace are:

requests.cpu: 2           # Soft limit on CPU usage
requests.memory: "6Gi"    # Soft limit on memory usage
limits.cpu: 2             # Hard limit on CPU usage
limits.memory: "8Gi"      # Hard limit on memory usage
pods: 4
services: 1
services.nodeport: 0      # Nodeport services are not allowed
replicationcontrollers: 1
secrets: 10
configmaps: 10
persistentvolumeclaims: 3

Quota increases

It is possible to request a quota increase if you can demonstrate your tool's need for more resources than the default namespace quota allows. Instructions and a template link for creating a quota request can be found at Toolforge (Quota requests) in Phabricator. Please read all the instructions there before submitting your request.

Container images

The Toolforge Kubernetes cluster is restricted to loading Docker images published at (see Portal:Toolforge/Admin/Kubernetes#Docker Images for more information). These images are built using the Dockerfiles in the operations/docker-images/toollabs-images git repository.

Available container types

The webservice command has an optional type argument that allows you to choose which Docker container to run your Tool in.

Currently provided types:

  • golang (go v1.11.5; deprecated)
  • golang111 (go v1.11.6)
  • jdk11 (openjdk 11.0.5)
  • jdk8 (openjdk 1.8.0_232; deprecated)
  • node10 (nodejs v10.15.2)
  • nodejs (nodejs v6.11.0; deprecated)
  • php5.6 (PHP 5.6.33; deprecated)
  • php7.2 (PHP 7.2.24; deprecated)
  • php7.3 (PHP 7.3.11)
  • python (Python 3.4.2; deprecated)
  • python2 (Python 2.7.9; deprecated)
  • python3.5 (Python 3.5.3; deprecated)
  • python3.7 (Python 3.7.3)
  • ruby2 (Ruby 2.1.5p273; deprecated)
  • ruby25 (Ruby 2.5.5p157)
  • tcl (TCL 8.6)

For example to start a webservice using a php7.3 container, run:

webservice --backend=kubernetes php7.3 start

A complete list of images is available from the docker-registry tool which provides a pretty frontend for browsing the Docker registry catalog.

As of Feb 2018, we don't support mixed runtime containers. This may change in the future. Also, we don't support "bring your own container" on our kubernetes (yet!). And there is no mechanism for a user to install system packages inside of a container.


PHP uses lighttpd as a webserver, and looks for files in ~/public_html/.

PHP versions & packages

There are three versions of PHP available, PHP 7.3 (on Debian Buster), PHP 7.2 (on Debian Stretch), and the legacy PHP 5.6 (on Debian Jessie).

You can view the installed PHP extensions on the phpinfo tool. This should match the PHP related packages installed on GridEngine exec nodes. Additional packages can be added on request by creating a Phabricator task tagged with #toolforge-software. Software that is not packaged by Debian upstream is less likely to be added due to security and maintenance concerns.

PHP Upgrade

To upgrade from PHP 5.6 to PHP 7.3, run the following two commands:

$ webservice stop
$ webservice --backend=kubernetes php7.3 start

To switch back:

$ webservice stop
$ webservice --backend=kubernetes php5.6 start

Running Locally

You may run the container on your local computer (not on Toolforge servers) by executing a command like this:

$ docker run --name toolforge -p 8888:80 -v "${PWD}:/var/www/html:cached" -d sh -c "lighty-enable-mod fastcgi-php && lighttpd -D -f /etc/lighttpd/lighttpd.conf"

Then the tool will be available at http://localhost:8888


The container images for Node.js, such as currently come with a current version of Node.js LTS from Wikimedia APT (as of September 2018, this is Node.js 6). This is the same version used by Wikimedia Foundation in production and for continuous integration.

Broken npm

Given npm is not suitable for use in Wikimedia production, the version of Node.js provided by Wikimedia APT is compiled without npm. (Unlike the official Node.js distribution.) And because there is no use for npm in Wikimedia production, there is no "npm" Debian package maintained in Wikimedia APT. The result is that the only "npm" Debian package available is the one from upstream Debian, which is npm 1.4 which was originally bundled in 2014 with Node 0.10 (debian/npm, debian/nodejs). This version is EOL and is incompatible with most packages on the registry. To update it within your container, follow these steps

# Step 1: Start a shell in your Node.js pod (see "Shell" section below)
tool@tools-login$ kubectl exec -it podname-123-aaa -- /bin/bash

# Step 2: Create $HOME/bin and ensure it is in your PATH
podname:/data/project/tool$ mkdir bin/
podname:/data/project/tool$ export PATH="${HOME}/bin:${PATH}"
# To avoid having to re-export PATH every time you use your tool, add the export command to your .bashrc file!

# Step 3: Use npm to install 'npm'
podname:/data/project/tool$ npm install npm
# This installs the current version of npm at node_modules/.bin/npm

# Step 4: Create a symlink in $HOME/bin
podname:/data/project/tool$ ln -s $HOME/node_modules/.bin/npm $HOME/bin/npm
# Close the shell and create a new shell (to initialise PATH)
podname:/data/project/tool$ exit
tool@tools-login$ kubectl exec -it podname-123-aaa -- /bin/bash

# Step 5: Verify that you now use a current npm instead of npm 1.4
podname:/data/project/tool$ npm --version


"failed to create new OS thread" from kubectl

If kubectl get pods or a similar command fails with the error message runtime: failed to create new OS thread (have 12 already; errno=11), use GOMAXPROCS=1 kubectl ... to reduce the number of resources that kubectl requests from the operating system.

Get a shell inside a running Pod

Kubectl can be used to open a shell inside a running Pod: $ kubectl exec -it $NAME_OF_POD -- /bin/bash

See Get a Shell to a Running Container at for more information.

Communication and support

We communicate and provide support through several primary channels. Please reach out with questions and to join the conversation.

Communicate with us
Connect Best for
Phabricator Workboard #Cloud-Services Task tracking and bug reporting
IRC Channel #wikimedia-cloud connect General discussion and support
Mailing List cloud@ Information about ongoing initiatives, general discussion and support
Announcement emails cloud-announce@ Information about critical changes (all messages mirrored to cloud@)
News wiki page News Information about major near-term plans
Cloud Services Blog Clouds & Unicorns Learning more details about some of our work
Wikimedia Technology Blog News and stories from the Wikimedia technical movement

See also