FAQ

General Questions

Can I use DevStack for production?

DevStack is targeted at developers and CI systems to use the raw upstream code. It makes many choices that are not appropriate for production systems.

Your best choice is probably to choose a distribution of OpenStack.

Can I use DevStack as a development environment?

Sure, you can. That said, there are a couple of things you should note before doing so:

  • DevStack makes a lot of configuration changes to your system and should not be run in your main development environment.

  • All the repositories that DevStack clones when deploying are considered volatile by default and thus are subject to hard resets. This is necessary to keep you in sync with the latest upstream, which is what you want in a CI situation, but it can result in branches being overwritten and files being removed.

    The corollary of this is that if you are working on a specific project, using the DevStack project repository (defaulted to /opt/stack/<project>) as the single master repository for storing all your work is not recommended. This behavior can be overridden by setting the RECLONE config option to no. Alternatively, you can avoid running stack.sh to redeploy by restarting services manually. In any case, you should generally ensure work in progress is pushed to Gerrit or otherwise backed up before running stack.sh.

  • If you use DevStack within a VM, you may wish to mount a local OpenStack directory, such as ~/src/openstack, inside the VM and configure DevStack to use this as the clone location using the {PROJECT}_REPO config variables. For example, assuming you’re using Vagrant and sharing your home directory, you should place the following in local.conf:

    NEUTRON_REPO=/home/vagrant/src/neutron
    NOVA_REPO=/home/vagrant/src/nova
    KEYSTONE_REPO=/home/vagrant/src/keystone
    GLANCE_REPO=/home/vagrant/src/glance
    SWIFT_REPO=/home/vagrant/src/swift
    HORIZON_REPO=/home/vagrant/src/horizon
    CINDER_REPO=/home/vagrant/src/cinder
    HEAT_REPO=/home/vagrant/src/heat
    TEMPEST_REPO=/home/vagrant/src/tempest
    HEATCLIENT_REPO=/home/vagrant/src/python-heatclient
    GLANCECLIENT_REPO=/home/vagrant/src/python-glanceclient
    NOVACLIENT_REPO=/home/vagrant/src/python-novaclient
    NEUTRONCLIENT_REPO=/home/vagrant/src/python-neutronclient
    OPENSTACKCLIENT_REPO=/home/vagrant/src/python-openstackclient
    HEAT_CFNTOOLS_REPO=/home/vagrant/src/heat-cfntools
    HEAT_TEMPLATES_REPO=/home/vagrant/src/heat-templates
    NEUTRON_FWAAS_REPO=/home/vagrant/src/neutron-fwaas
    # ...
    

Why a shell script, why not chef/puppet/…

The script is meant to be read by humans (as well as ran by computers); it is the primary documentation after all. Using a recipe system requires everyone to agree and understand chef or puppet.

I’d like to help!

That isn’t a question, but please do! The source for DevStack is at opendev.org and bug reports go to LaunchPad. Contributions follow the usual process as described in the developer guide. This Sphinx documentation is housed in the doc directory.

Why not use packages?

Unlike packages, DevStack leaves your cloud ready to develop - checkouts of the code and services running locally under systemd, making it easy to hack on and test new patches. However, many people are doing the hard work of packaging and recipes for production deployments.

Why isn’t $MY_FAVORITE_DISTRO supported?

DevStack is meant for developers and those who want to see how OpenStack really works. DevStack is known to run on the distro/release combinations listed in README.md. DevStack is only supported on releases other than those documented in README.md on a best-effort basis.

Are there any differences between Ubuntu and CentOS/Fedora support?

Both should work well and are tested by DevStack CI.

Why can’t I use another shell?

DevStack now uses some specific bash-ism that require Bash 4, such as associative arrays. Simple compatibility patches have been accepted in the past when they are not complex, at this point no additional compatibility patches will be considered except for shells matching the array functionality as it is very ingrained in the repo and project management.

Can I test on OS/X?

Some people have success with bash 4 installed via homebrew to keep running tests on OS/X.

Can I at least source openrc with zsh?

People have reported success with a special function to run openrc through bash for this

function sourceopenrc {
    pushd ~/devstack >/dev/null
    eval $(bash -c ". openrc $1 $2 >/dev/null;env|sed -n '/OS_/ { s/^/export /;p}'")
    popd >/dev/null
}

Operation and Configuration

How can I document the environment that DevStack is using?

DevStack includes a script (tools/info.sh) that gathers the versions of the relevant installed apt packages, pip packages and git repos. This is a good way to verify what Python modules are installed.

How do I turn off a service that is enabled by default?

Services can be turned off by adding disable_service xxx to local.conf (using c-vol in this example):

disable_service c-vol

How do I run a specific OpenStack release?

DevStack master tracks the upstream master of all the projects. If you would like to run a stable branch of OpenStack, you should use the corresponding stable branch of DevStack as well. For instance the stable/ocata version of DevStack will already default to all the projects running at stable/ocata levels.

Note: it’s also possible to manually adjust the *_BRANCH variables further if you would like to test specific milestones, or even custom out of tree branches. This is done with entries like the following in your local.conf

[[local|localrc]]
GLANCE_BRANCH=11.0.0.0rc1
NOVA_BRANCH=12.0.0.0.rc1

Upstream DevStack is only tested with master and stable branches. Setting custom BRANCH definitions is not guaranteed to produce working results.

What can I do about RabbitMQ not wanting to start on my fresh new VM?

This is often caused by erlang not being happy with the hostname resolving to a reachable IP address. Make sure your hostname resolves to a working IP address; setting it to 127.0.0.1 in /etc/hosts is often good enough for a single-node installation. And in an extreme case, use clean.sh to eradicate it and try again.

Why are my configuration changes ignored?

You may have run into the package prerequisite installation timeout. tools/install_prereqs.sh has a timer that skips the package installation checks if it was run within the last PREREQ_RERUN_HOURS hours (default is 2). To override this, set FORCE_PREREQ=1 and the package checks will never be skipped.

Miscellaneous

tools/fixup_stuff.sh is broken and shouldn’t ‘fix’ just one version of packages.

Stuff in there is to correct problems in an environment that need to be fixed elsewhere or may/will be fixed in a future release. In the case of httplib2 and prettytable specific problems with specific versions are being worked around. If later releases have those problems than we’ll add them to the script. Knowing about the broken future releases is valuable rather than polling to see if it has been fixed.