This repository contains an easy to replicate recipe to spin up a Kopano demo environment, which can (through modification of .env and possibly docker-compose.yml/docker-compose.override.yml) also be used for production environments.
- make sure that you are running at least Docker 17.06.0 and Docker Compose 1.19.0.
- clone this repository to your local disk
- run
setup.sh:- this script will ask you a few questions about your environment
- If you are just interested in the demo environment you can accept the default values by pressing
Enteron each question
- now run
docker-compose upand you will see how the remaining Docker images are pulled and then everything is started - after startup has succeeded you can access the Kopano WebApp by going to
https://kopano.demo/webapp - there are already some users created in the demo LDAP
- these users all have a password that is identical to the username, e.g. the password for
user1useruser1
- these users all have a password that is identical to the username, e.g. the password for
- to build own containers at least Docker 19.03 is required
- this is due to the usage of build-time secrets
If you want to get an impression how the containers interact/relate with each other have a look at the architecture description.
Note: There have been reports about the LDAP demo not starting up on MacOS. It is recommended to use a Linux OS if you want to use the bundled LDAP image.
The docker-compose.yml file by default pulls Docker containers from for example https://hub.docker.com/r/zokradonh/kopano_core/ and https://hub.docker.com/r/zokradonh/kopano_webapp/. These images are based on the Kopano nightly builds and will contain the latest version available from the time the image was built.
If you are running into a problem please include the following issues in the description of your report:
- the error message produced when running
docker-compose up- or the output of
docker-compose ps
- or the output of
- for failed containers the output of
docker-compose logs $containername - the contents of your
.env - either the output of
docker-compose config(only useful whendocker-compose upsucceeds) or yourdocker-compose.yml - the output of
docker -vanddocker-compose -v
Please make sure to remove sensitive data (for example your real hostname or passwords for accounts) when posting these information publicly.
If you have problems or questions about Kopano in general then either get in contact with the Kopano support (if you have a valid subscription) or start a topic on the Kopano forum.
The used docker-compose.yml is part of the git versioning. Which mean all changes in this repository will also be applied to your local data once you pull in the latest version. If you need to configure additional env variables, this can now be done in the additional env files (more details further below, for example for kopano_server this file is called kopano_server.env). If you only want to run a subset of containers it is recommended to create a copy of docker-compose.yml and specify your copy when running. e.g. like docker-compose -f my-setup.yml up -d.
Yes, indeed. You could for example use this to easily try out newer Kopano WebApp or Z-Push releases, without touching your production environment. Through the zokradonh/kopano_core image you could even try out newer version of e.g. kopano-gateway without jumping into a dependency mess in your production environment.
And last but not least this project also offers a zokradonh/kopano_utils image to easily run tools such as kopano-backup, kopano-migration-pst, kopano-migration-imap and all the other utilities that are bundles with Kopano. See below to see how to run zokradonh/kopano_utils.
If you want to modify some of the values from the setup.sh run you can simply edit .env in your favorite editor. Repeated runs of setup.sh will neither modify docker-compose.yml nor .env. In the ´.env´ file you will also find some given defaults like LDAP query filters and the local ports for the reverse proxy.
Additionally a dedicated env file is created for each container (at least where that would make sense). The env file has the container name as part of the file name. For example for the kopano_server container the filename is named kopano_server.env. These additional env files are auto created when running setup.sh.
Any additional configuration should be done through environment variables and not done in the actual container. The images working with configuration files (e.g. kopano_core, kopano_webapp, kopano_meet) have a mechanism built in to translate env variables into configuration files. For services that can directly work with env variables (e.g. kopano_konnect, kopano_kwmserver) these can be specified directly. Please check the individual README.md files for further instructions.
The compose file itself is part of the git repository and should not be edited directly. Instead a docker-compose.override.yml file (will be ignored by git) can be created to override and extend the default one. See https://docs.docker.com/compose/extends/ for more information.
This project uses the COMPOSE_FILE environment variable to allow users to override the ports exposed by each container (for example when using a different front facing proxy). When using a docker-compose.override.yml file make sure this is included in this variable in the .env file. For example like this:
COMPOSE_FILE=docker-compose.yml:docker-compose.portmapping.yml:docker-compose.override.ymlThis project switched to specifying COMPOSE_FILE in .env to allow users to easily disable individual ports exposed. Please rerun setup.sh to add this variable to your .env file or add it manually.
In this repository you can also find a Makefile that automates the process of building newer images.
You can easily rebuild all images based on the currently available Kopano version by running make build-all. To just rebuild a certain image you can also run make build-core or make build-webapp. Please check the Makefile to see other possible targets. (depending on your environment you may also be able to autocomplete with the Tab key)
To be able to easily go back to a previous version you can also "tag" you Docker images by running e.g. make tag-core.
There are certain tasks within Kopano that either need to be executed once (like creating the public store when starting a new environment for the first time) or on a regular base (like syncing the internal user list with and external LDAP tree). For convenience this project includes a scheduler container that will take care of this and that can be dynamically configured through env variables.
Please check the README.md of the scheduler image for further instructions.
Instead of using the internal scheduler one can also just use an existing scheduler (cron on the docker host for example) to execute these tasks.
This project also makes it possible to build Docker images based on the official Kopano releases. For this the following section needs to be modified in .env:
# Docker Repository to push to/pull from
docker_repo=zokradonh
COMPOSE_PROJECT_NAME=kopano
COMPOSE_FILE=docker-compose.yml:docker-compose.ports.yml:docker-compose.db.yml:docker-compose.ldap.yml:docker-compose.mail.yml
# Modify below to build a different version, than the Kopano nightly release
# credentials for repositories are handled through a file called apt_auth.conf (which will be created through setup.sh or Makefile)
#KOPANO_CORE_REPOSITORY_URL=https://download.kopano.io/supported/core:/8.7/Debian_10/
#KOPANO_MEET_REPOSITORY_URL=https://download.kopano.io/supported/meet:/final/Debian_10/
#KOPANO_WEBAPP_REPOSITORY_URL=https://download.kopano.io/supported/webapp:/final/Debian_10/
#KOPANO_WEBAPP_FILES_REPOSITORY_URL=https://download.kopano.io/supported/files:/final/Debian_10/
#KOPANO_WEBAPP_MDM_REPOSITORY_URL=https://download.kopano.io/supported/mdm:/final/Debian_10/
#KOPANO_WEBAPP_SMIME_REPOSITORY_URL=https://download.kopano.io/supported/smime:/final/Debian_10/
#KOPANO_ZPUSH_REPOSITORY_URL=https://download.kopano.io/zhub/z-push:/final/Debian_10/
#DOWNLOAD_COMMUNITY_PACKAGES=0The credentials for the Kopano package repositories can either be defined through the url itself, e.g. like https://serial:REPLACE-ME@download.kopano.io/supported/core:/final/Debian_10/ or through an apt_auth.conf file. Using apt_auth.conf is preferred, since it does not "leak" credentials into the final image.
With the above lines uncommented and credentials in place running make build-all will rebuild the images based on the latest available Kopano release (don't forget to make tag-core and make tag-webapp your images after building them).
If you are running a private Docker Registry then you have to change docker_repo to reference your internal registry. Afterward you can run for example make publish-core to push the image to your registry.
WARNING
When storing the credentials in the url the built image will include your subscription key! Do not push this image to any public registry like e.g. https://hub.docker.com!
Please check the contributing information.
What if I want to use a different front facing proxy than the one in docker-compose? Or just some part of the compose file?
While using kweb is recommended, this is of course possible.
Please check the individual web containers (kDAV, WebApp and Z-Push for individual instructions).
The exposed ports of each container are defined in docker-compose.ports.yml. If you do not want to expose some of the containers to the network, it is recommended to copy this file to docker-compose.override.ymland just remove all entries that you do not want to have exposed.
I want to use these Docker images outside of an evaluation environment. What do I need to adjust to make this possible?
To get a quick impression of Kopano this git repository bundles a locally build LDAP image with some example users. When using the docker-compose.yml in a production environment make sure to:
- switch to the non-demo ldap tree or completely remove the local LDAP from the compose file
- adapt LDAP queries in .env to match you actual LDAP server and users
- all additional configuration of the Kopano components should be specified in the compose file/the env file/an override and not within the running container
- make sure that there is a unique machine-id for your deployment
- the default setup mounts the file from the host, if your host is running multiple installations of Kopano make sure to generate a unique value for each installation.
Yes, that is certainly a possibility. Within the examples/ directory you can find some ready to run examples that can be run in the following way:
docker-compose -f examples/webapp.yml up -d
- Start
docker-compose.ymlfile in the background:docker-compose up -d - Get a status overview of the running containers:
docker-compose ps - Stop compose running in the background:
docker-compose stop - Destroy local containers and network interfaces:
docker-compose down - Destroy volumes as well (will completely reset the containers, deletes all data):
docker-compose down -v - Run commands in a running container:
docker-compose exec kopano_server kopano-admin -l - Get logs of a in the background running container:
docker-compose logs -f kopano_server - Run a
kopano-backup:docker run --rm -it -v /var/run/kopano/:/var/run/kopano -v $(pwd):/kopano/path zokradonh/kopano_utils kopano-backup- Same command but getting volumes from the existing
kopano_servercontainer:docker run --rm -it --volumes-from kopano_server -v /root/kopano-backup:/kopano/path zokradonh/kopano_utils kopano-backup -h
- Same command but getting volumes from the existing
- Get a shell in a new container to (for example) run
kopano-migration-pst:docker run --rm -it -v /var/run/kopano/:/var/run/kopano -v $(pwd):/kopano/path zokradonh/kopano_utils(to directly run kopano-migration-pst just append it to the command)
This project includes a configuration file for Vagrant to easily try kopano-docker locally. All that is required is Vagrant itself and Virtualbox.
Steps to start kopano-docker in Vagrant:
# run setup.sh
$ ./setup.sh
# provision virtual machine
$ vagrant up
# alternatively "vagrant up --provider hyperv" when running on Windows
# in case you want to connect into the machine
$ vagrant sshAfter the machine has started it will be reachable from the local system through the IP 10.16.73.20, please make sure that your chosen hostname resolves to this IP. The project files are mounted to /vagrant in the machine. To interact with the containers just change into this directory first.
The example docker-compose.yml uses the following components for the MTA (mail delivery, including anti-spam & anti-virus) and openLDAP. Please consult their documentation for further configuration advice.
- https://github.com/tomav/docker-mailserver/
- https://github.com/osixia/docker-openldap
- https://github.com/osixia/docker-phpLDAPadmin
The following (blog) articles have been written about this project: