1.3.0 RC1

This is a release candidate for version 1.3.0 of Docker Compose. To try it out, run the following commands:

curl -L https://github.com/docker/compose/releases/download/1.3.0rc1/docker-compose-`uname -s`-`uname -m` > /usr/local/bin/docker-compose
chmod +x /usr/local/bin/docker-compose

Or pip install -U docker-compose==1.3.0rc1 if you prefer Pip.

Now, two important notes:

• This release contains breaking changes, and you will need to either remove or migrate your existing containers before running your app - see the upgrading section of the install docs for details.
• Compose now requires Docker 1.6.0 or later.

We've done a lot of work in this release to remove hacks and make Compose more stable:

• Compose now uses Docker labels, rather than container names, to keep track of containers. This is both cleaner and more performant.
• Compose no longer uses "intermediate containers" when recreating containers for a service. This makes docker-compose up less complex and more resilient to failure.

There are some new features:

• docker-compose up has an experimental new behaviour: it will only recreate containers for services whose configuration has changed in docker-compose.yml. This will eventually become the default, but for now you can take it for a spin:

  $ docker-compose up --x-smart-recreate
  

• When invoked in a subdirectory of a project, docker-compose will now climb up through parent directories until it finds a docker-compose.yml.

Several new configuration keys have been added to docker-compose.yml:

• dockerfile, like docker build --file, lets you specify an alternate Dockerfile to use with build.
• labels, like docker run --labels, lets you add custom metadata to containers.
• extra_hosts, like docker run --add-host, lets you add entries to a container's /etc/hosts file.
• pid: host, like docker run --pid=host, lets you reuse the same PID namespace as the host machine.
• cpuset, like docker run --cpuset-cpus, lets you specify which CPUs to allow execution in.
• read_only, like docker run --read-only, lets you mount a container's filesystem as read-only.
• security_opt, like docker run --security-opt, lets you specify security options.
• log_driver, like docker run --log-driver, lets you specify a log driver.

Many bugs have been fixed, including the following:

• The output of docker-compose run was sometimes truncated, especially when running under Jenkins.
• A service's volumes would sometimes not update after volume configuration was changed in docker-compose.yml.
• Authenticating against third-party registries would sometimes fail.
• docker-compose run --rm would fail to remove the container if the service had a restart policy in place.

Thanks @ahromis, @albers, @aleksandr-vin, @antoineco, @ccverak, @chernjie, @dnephin, @josephpage, @KyleJamesWalker, @lsowen, @mchasal, @sdake, @sherter, @stephenlawrence, @turtlemonvh, @vdemeester, @xuxinkun and @zwily!

1.2.0

To install Compose 1.2.0, run the following commands:

curl -L https://github.com/docker/compose/releases/download/1.2.0/docker-compose-`uname -s`-`uname -m` > /usr/local/bin/docker-compose
chmod +x /usr/local/bin/docker-compose

Optionally, you can also install command completion for the bash shell.

Here's what's new:

• docker-compose.yml now supports an extends option, which enables a service to inherit configuration from another service in another configuration file. This is really good for sharing common configuration between apps, or for configuring the same app for different environments. Here's the documentation.
• When using Compose with a Swarm cluster, containers that depend on one another will be co-scheduled on the same node. This means that most Compose apps will now work out of the box, as long as they don't use build.
• Repeated invocations of docker-compose up when using Compose with a Swarm cluster now work reliably.
• Directories passed to build, filenames passed to env_file and volume host paths passed to volumes are now treated as relative to the directory of the configuration file, not the directory that docker-compose is being run in. In the majority of cases, those are the same, but if you use the -f|--file argument to specify a configuration file in another directory, this is a breaking change.
• A service can now share another service's network namespace with net: container:<service>.
• volumes_from and net: container:<service> entries are taken into account when resolving dependencies, so docker-compose up <service> will correctly start all dependencies of <service>.
• docker-compose run now accepts a --user argument to specify a user to run the command as, just like docker run.
• The up, stop and restart commands now accept a --timeout (or -t) argument to specify how long to wait when attempting to gracefully stop containers, just like docker stop.
• docker-compose rm now accepts -f as a shorthand for --force, just like docker rm.

Thanks, @abesto, @albers, @alunduil, @dnephin, @funkyfuture, @gilclark, @IanVS, @KingsleyKelly, @knutwalker, @thaJeztah and @vmalloc!

1.2.0 RC4

This is a release candidate for Compose 1.2.0. To try it out, run the following commands:

curl -L https://github.com/docker/compose/releases/download/1.2.0rc4/docker-compose-`uname -s`-`uname -m` > /usr/local/bin/docker-compose
chmod +x /usr/local/bin/docker-compose

Optionally, you can also install command completion for the bash shell.

On top of the changes listed below for RC1-RC3, the following bugs have been fixed:

• Environment variables and the user's home directory (~) were not expanding properly in volumes host paths.

Here are the changes for RC3:

• When copying a service's configuration with extends, image and build could come into conflict, resulting in an error, as it makes no sense to have both defined. Each now overwrites the other: if a service with image defined is extended and build is added, the image entry will be removed.
• When copying a service's configuration with extends, if both services defined a multi-value option such as ports or dns, the original value would be completely discarded. They are now concatenated instead.
• When a relative path is supplied to build, it is treated as relative to the directory of the configuration file, not the directory that docker-compose is being run in. In the majority of cases, those are the same, but if you use the -f|--file argument to specify a configuration file in another directory, this is a breaking change.

Here are the changes for RC2:

• A bug has been fixed where containers were being created with blank entries for "Dns" and "DnsSearch", causing DNS lookups from within a container to fail.

Here are the changes for RC1:

• docker-compose.yml now supports an extends option, which enables a service to inherit configuration from another service in another configuration file. This is really good for sharing common configuration between apps, or for configuring the same app for different environments. Here's the documentation.
• When using Compose with a Swarm cluster, containers that depend on one another will be co-scheduled on the same node. This means that most Compose apps will now work out of the box, as long as they don't use build.
• Repeated invocations of docker-compose up when using Compose with a Swarm cluster now work reliably.
• Filenames in env_file and volume host paths in volumes are now treated as relative to the directory of the configuration file, not the directory that docker-compose is being run in. In the majority of cases, those are the same, but if you use the -f|--file argument to specify a configuration file in another directory, this is a breaking change.
• A service can now share another service's network namespace with net: container:<service>.
• volumes_from and net: container:<service> entries are taken into account when resolving dependencies, so docker-compose up <service> will correctly start all dependencies of <service>.
• Problems with authentication when using images from third-party registries have been fixed.
• docker-compose run now accepts a --user argument to specify a user to run the command as, just like docker run.
• The up, stop and restart commands now accept a --timeout (or -t) argument to specify how long to wait when attempting to gracefully stop containers, just like docker stop.
• docker-compose rm now accepts -f as a shorthand for --force, just like docker rm.

Thanks, @abesto, @albers, @alunduil, @dnephin, @funkyfuture, @gilclark, @IanVS, @KingsleyKelly, @knutwalker, @moysesb, @thaJeztah and @vmalloc!

1.2.0 RC3

This is a release candidate for Compose 1.2.0. To try it out, run the following commands:

curl -L https://github.com/docker/compose/releases/download/1.2.0rc3/docker-compose-`uname -s`-`uname -m` > /usr/local/bin/docker-compose
chmod +x /usr/local/bin/docker-compose

Optionally, you can also install command completion for the bash shell.

On top of the changes listed below for RC1 and RC2, the following bugs have been fixed:

• When copying a service's configuration with extends, image and build could come into conflict, resulting in an error, as it makes no sense to have both defined. Each now overwrites the other: if a service with image defined is extended and build is added, the image entry will be removed.
• When copying a service's configuration with extends, if both services defined a multi-value option such as ports or dns, the original value would be completely discarded. They are now concatenated instead.
• When a relative path is supplied to build, it is treated as relative to the directory of the configuration file, not the directory that docker-compose is being run in. In the majority of cases, those are the same, but if you use the -f|--file argument to specify a configuration file in another directory, this is a breaking change.

Here are the changes for RC2:

• A bug has been fixed where containers were being created with blank entries for "Dns" and "DnsSearch", causing DNS lookups from within a container to fail.

Here are the changes for RC1:

• docker-compose.yml now supports an extends option, which enables a service to inherit configuration from another service in another configuration file. This is really good for sharing common configuration between apps, or for configuring the same app for different environments. Here's the documentation.
• When using Compose with a Swarm cluster, containers that depend on one another will be co-scheduled on the same node. This means that most Compose apps will now work out of the box, as long as they don't use build.
• Repeated invocations of docker-compose up when using Compose with a Swarm cluster now work reliably.
• Filenames in env_file and volume host paths in volumes are now treated as relative to the directory of the configuration file, not the directory that docker-compose is being run in. In the majority of cases, those are the same, but if you use the -f|--file argument to specify a configuration file in another directory, this is a breaking change.
• A service can now share another service's network namespace with net: container:<service>.
• volumes_from and net: container:<service> entries are taken into account when resolving dependencies, so docker-compose up <service> will correctly start all dependencies of <service>.
• Problems with authentication when using images from third-party registries have been fixed.
• docker-compose run now accepts a --user argument to specify a user to run the command as, just like docker run.
• The up, stop and restart commands now accept a --timeout (or -t) argument to specify how long to wait when attempting to gracefully stop containers, just like docker stop.
• docker-compose rm now accepts -f as a shorthand for --force, just like docker rm.

Thanks, @abesto, @albers, @alunduil, @dnephin, @funkyfuture, @gilclark, @IanVS, @KingsleyKelly, @knutwalker, @moysesb, @thaJeztah and @vmalloc!

1.2.0 RC2

This is a release candidate for Compose 1.2.0. To try it out, run the following commands:

curl -L https://github.com/docker/compose/releases/download/1.2.0rc2/docker-compose-`uname -s`-`uname -m` > /usr/local/bin/docker-compose
chmod +x /usr/local/bin/docker-compose

Optionally, you can also install command completion for the bash shell.

• On top of the changes listed below for RC1, a bug has been fixed where containers were being created with blank entries for "Dns" and "DnsSearch", causing DNS lookups from within a container to fail.

Here are the changes for RC1:

• docker-compose.yml now supports an extends option, which enables a service to inherit configuration from another service in another configuration file. This is really good for sharing common configuration between apps, or for configuring the same app for different environments. Here's the documentation.
• When using Compose with a Swarm cluster, containers that depend on one another will be co-scheduled on the same node. This means that most Compose apps will now work out of the box, as long as they don't use build.
• Repeated invocations of docker-compose up when using Compose with a Swarm cluster now work reliably.
• Filenames in env_file and volume host paths in volumes are now treated as relative to the directory of the configuration file, not the directory that docker-compose is being run in. In the majority of cases, those are the same, but if you use the -f|--file argument to specify a configuration file in another directory, this is a breaking change.
• A service can now share another service's network namespace with net: container:<service>.
• volumes_from and net: container:<service> entries are taken into account when resolving dependencies, so docker-compose up <service> will correctly start all dependencies of <service>.
• Problems with authentication when using images from third-party registries have been fixed.
• docker-compose run now accepts a --user argument to specify a user to run the command as, just like docker run.
• The up, stop and restart commands now accept a --timeout (or -t) argument to specify how long to wait when attempting to gracefully stop containers, just like docker stop.
• docker-compose rm now accepts -f as a shorthand for --force, just like docker rm.

Thanks, @abesto, @albers, @alunduil, @dnephin, @funkyfuture, @gilclark, @IanVS, @KingsleyKelly, @knutwalker, @thaJeztah and @vmalloc!

1.2.0 RC1

This is a release candidate for Compose 1.2.0. To try it out, run the following commands:

curl -L https://github.com/docker/compose/releases/download/1.2.0rc1/docker-compose-`uname -s`-`uname -m` > /usr/local/bin/docker-compose
chmod +x /usr/local/bin/docker-compose

Optionally, you can also install command completion for the bash shell.

Here's what's new:

• docker-compose.yml now supports an extends option, which enables a service to inherit configuration from another service in another configuration file. This is really good for sharing common configuration between apps, or for configuring the same app for different environments. Here's the documentation.
• When using Compose with a Swarm cluster, containers that depend on one another will be co-scheduled on the same node. This means that most Compose apps will now work out of the box, as long as they don't use build.
• Repeated invocations of docker-compose up when using Compose with a Swarm cluster now work reliably.
• Filenames in env_file and volume host paths in volumes are now treated as relative to the directory of the configuration file, not the directory that docker-compose is being run in. In the majority of cases, those are the same, but if you use the -f|--file argument to specify a configuration file in another directory, this is a breaking change.
• A service can now share another service's network namespace with net: container:<service>.
• volumes_from and net: container:<service> entries are taken into account when resolving dependencies, so docker-compose up <service> will correctly start all dependencies of <service>.
• Problems with authentication when using images from third-party registries have been fixed.
• docker-compose run now accepts a --user argument to specify a user to run the command as, just like docker run.
• The up, stop and restart commands now accept a --timeout (or -t) argument to specify how long to wait when attempting to gracefully stop containers, just like docker stop.
• docker-compose rm now accepts -f as a shorthand for --force, just like docker rm.

Thanks, @abesto, @albers, @alunduil, @dnephin, @funkyfuture, @gilclark, @IanVS, @KingsleyKelly, @knutwalker, @thaJeztah and @vmalloc!

1.1.0

Fig has been renamed to Docker Compose, or just Compose for short. This has several implications for you:

• The command you type is now docker-compose, not fig.
• You should rename your fig.yml to docker-compose.yml.
• If you’re installing via PyPi, the package is now docker-compose, so install it with pip install docker-compose.

To install Compose, run the following commands:

curl -L https://github.com/docker/compose/releases/download/1.1.0/docker-compose-`uname -s`-`uname -m` > /usr/local/bin/docker-compose
chmod +x /usr/local/bin/docker-compose

Besides that, there’s a lot of new stuff in this release:

• docker-compose run now has a --service-ports flag for exposing ports on the given service. This is useful for e.g. running your webapp with an interactive debugger.
• You can now link to containers outside your app with the external_links option in docker-compose.yml.
• You can now prevent docker-compose up from automatically building images with the --no-build option. This will make fewer API calls and run faster.
• If you don’t specify a tag when using the image key, Compose will default to the latest tag, rather than pulling all tags.
• docker-compose kill now supports the -s flag, allowing you to specify the exact signal you want to send to a service’s containers.
• docker-compose.yml now has an env_file key, analogous to docker run --env-file, letting you specify multiple environment variables in a separate file. This is great if you have a lot of them, or if you want to keep sensitive information out of version control.
• docker-compose.yml now supports the dns_search, cap_add, cap_drop, cpu_shares and restart options, analogous to docker run’s --dns-search, --cap-add, --cap-drop, --cpu-shares and --restart options.
• Compose now ships with Bash tab completion - see the installation and usage docs at https://github.com/docker/compose/blob/1.1.0/docs/completion.md
• We’ve made a few small changes to ensure that Compose will work with Swarm, Docker’s new clustering tool (https://github.com/docker/swarm). Eventually you'll be able to point Compose at a Swarm cluster instead of a standalone Docker host and it’ll run your containers on the cluster with no extra work from you. As Swarm is still developing, integration is rough and lots of Compose features don't work yet. Check the integration doc to see how it's progressing.
• A number of bugs have been fixed - see the milestone for details: https://github.com/docker/compose/issues?q=milestone%3A1.1.0+

Thanks @dnephin, @squebe, @jbalonso, @raulcd, @benlangfeld, @albers, @ggtools, @bersace, @dtenenba, @petercv, @drewkett, @TFenby, @paulRbr, @Aigeruth and @salehe!

1.1.0-rc2

Fig has been renamed to Docker Compose, or just Compose for short. This has several implications for you:

• The command you type is now docker-compose, not fig.
• You should rename your fig.yml to docker-compose.yml.
• The PyPi package is now docker-compose.

This is a release candidate for Docker Compose 1.1.0. To try it out, install via curl:

curl -L https://github.com/docker/fig/releases/download/1.1.0-rc2/docker-compose-`uname -s`-`uname -m` > /usr/local/bin/docker-compose; chmod +x /usr/local/bin/docker-compose

(If you install via PyPi, use pip install docker-compose==1.1.0-rc2.)

On top of the changelog for 1.1.0-rc1 (see below), the following bugs have been fixed:

• When an environment variables file specified with env_file doesn't exist, Compose was showing a stack trace instead of a helpful error.
• Configuration files using the old name (fig.yml) were not being read unless explicitly specified with docker-compose -f. Compose now reads them and prints a deprecation warning.
• Bash tab completion now reads fig.yml if it's present.

Here's the changelog for 1.1.0-rc1:

• We’ve made a few small changes to ensure that Compose will work with Swarm, Docker’s new clustering tool (https://github.com/docker/swarm). Eventually you'll be able to point Compose at a Swarm cluster instead of a standalone Docker host and it’ll run your containers on the cluster with no extra work from you. As Swarm is still developing, integration is rough and lots of Compose features don't work yet.
• docker-compose run now has a --service-ports flag for exposing ports on the given service. This is useful for e.g. running your webapp with an interactive debugger.
• You can now link to containers outside your app with the external_links option in docker-compose.yml.
• You can now prevent docker-compose up from automatically building images with the --no-build option. This will make fewer API calls and run faster.
• If you don’t specify a tag when using the image key, Compose will default to the latest tag, rather than pulling all tags.
• docker-compose kill now supports the -s flag, allowing you to specify the exact signal you want to send to a service’s containers.
• docker-compose.yml now has an env_file key, analogous to docker run --env-file, letting you specify multiple environment variables in a separate file. This is great if you have a lot of them, or if you want to keep sensitive information out of version control.
• docker-compose.yml now supports the dns_search, cap_add, cap_drop and restart options, analogous to docker run’s --dns-search, --cap-add, --cap-drop and --restart options.
• Compose now ships with Bash tab completion - see the installation and usage docs at https://github.com/docker/fig/blob/1.1.0-rc1/docs/completion.md
• A number of bugs have been fixed - see the milestone for details: https://github.com/docker/fig/issues?q=milestone%3A1.1.0+

1.1.0-rc1

Fig has been renamed to Docker Compose, or just Compose for short. This has several implications for you:

• The command you type is now docker-compose, not fig.
• You should rename your fig.yml to docker-compose.yml.
• The PyPi package is now docker-compose.

This is a release candidate for Docker Compose 1.1.0. To try it out, install via curl:

curl -L https://github.com/docker/fig/releases/download/1.1.0-rc1/docker-compose-`uname -s`-`uname -m` > /usr/local/bin/docker-compose; chmod +x /usr/local/bin/docker-compose

(If you install via PyPi, use pip install docker-compose==1.1.0-rc1.)

Besides the rename, there’s a lot of new stuff:

• We’ve made a few small changes to ensure that Compose will work with Swarm, Docker’s new clustering tool (https://github.com/docker/swarm). Eventually you'll be able to point Compose at a Swarm cluster instead of a standalone Docker host and it’ll run your containers on the cluster with no extra work from you. As Swarm is still developing, integration is rough and lots of Compose features don't work yet.
• docker-compose run now has a --service-ports flag for exposing ports on the given service. This is useful for e.g. running your webapp with an interactive debugger.
• You can now link to containers outside your app with the external_links option in docker-compose.yml.
• You can now prevent docker-compose up from automatically building images with the --no-build option. This will make fewer API calls and run faster.
• If you don’t specify a tag when using the image key, Compose will default to the latest tag, rather than pulling all tags.
• docker-compose kill now supports the -s flag, allowing you to specify the exact signal you want to send to a service’s containers.
• docker-compose.yml now has an env_file key, analogous to docker run --env-file, letting you specify multiple environment variables in a separate file. This is great if you have a lot of them, or if you want to keep sensitive information out of version control.
• docker-compose.yml now supports the dns_search, cap_add, cap_drop and restart options, analogous to docker run’s --dns-search, --cap-add, --cap-drop and --restart options.
• A number of bugs have been fixed - see the milestone for details: https://github.com/docker/fig/issues?q=milestone%3A1.1.0+

1.0.1

Install/upgrade instructions here.

• Added an --allow-insecure-ssl option to allow fig up, fig run and fig pull to pull from insecure registries.
• Fixed fig run not showing output in Jenkins.
• Fixed a bug where Fig couldn't build Dockerfiles with ADD statements pointing at URLs.