Compose (docker-compose) and Docker (docker) CLI
This page outlines commonly used docker-compose and docker commands used in GroundWork Monitor.
- Important! Mistakes can be made when typing Docker commands. You may want to consider a system backup.
- For additional reference see Compose command-line reference and Docker commands. You can also see this information by running docker-compose [SUBCOMMAND] --help from the command line
- Note docker-compose service names are the short names for the containers as listed in gw8/docker-compose.yml. , e.g., revproxy, cloudhub, groundwork, nagios. The .yml file should not be edited.
- You will need to be in your GroundWork Monitor 8 installation directory (e.g., /path-to/gw8) before executing commands.
- To create a shortcut for use when entering docker-compose commands you can create an alias. This enables you to use dc <command> versus the full docker-compose <command>. Just add an alias in the /home/gwos directory and bashrc file, e.g.,:
- Edit bashrc with the command:
- At the bottom of the file add the following:
- Save the file and try it out, with the dc command alone you should receive a list of docker compose commands:
sudo su - gwos
- Edit bashrc with the command:
List running containers
List all containers
List running GroundWork containers
|Start Stop Restart Containers
docker-compose up -d
Start GroundWork containers
Stop GroundWork containers
Restart GroundWork containers
|Start Stop Restart Individual Containers
Start individual container (service)
Stop individual container (service)
docker-compose restart <service>
Restart individual service
Force stop a container (service)
Execute a command running in a container
Display (and follow) a log file for a container
Display and follow log file (just add -f to above):
By default, access logging is disabled on all containers. However, there are Docker commands to enable them temporarily for diagnostic reasons. As logging is more expensive with the ELK Stack integrated, access to logs should not be left on during normal operation. Containers log independently, so any combination of the containers below can be enabled/disabled. Normally, revproxy access logging is sufficient to debug client/server communications.
Check Docker Access
Verify access to Docker
Verify access to Docker Compose
List all volumes
Show Docker version
Show Compose version
Cloud Hub Connections
This command will give you a line of output for each deployed connector instance, and will also show you the display name so you can disambiguate multiple deployments of the same connector type.
Configuring the Details
In many cases, no additional configuration beyond what is possible in the user interface will be needed. There are, however, some areas of the system that support customizations and tuning. If you are an advanced user, you will need to access these to get the most out of your system. To use these customizations and make tuning changes, you will need to edit some of the files the system uses to control the applications, settings, and parameters.
For example, we ship the NeDi application with no access to the database snapshot interface. If you want to use it, you must change the following line (un-comment) in the file /usr/local/groundwork/config/nedi/nedi.conf:
The file should then be saved, and the permissions should be left at -rw-rw-r-- 1 nedi nedi. That is, owned by the nedi user.
Such changes are preserved across upgrades.
Additions of files and deletions of files are NOT preserved, but will be reverted on upgrades. This ensures we can provide you with a secure, functional system with minimal instability.
While we are working on adding more ways to configure GroundWork Monitor to the user interface, we expect for the foreseeable future some aspects will require command line editing of configuration files, and possibly the restarting of containers. The following instructions are provided to enable you to make these changes safely and without having to master too many new commands.
Generally speaking, you will want to have your GroundWork Monitor 8 system running, and use a docker feature called docker exec to make changes in the containers. Don't try to go around the containers and access the files directly on the file system. The containers are there for your protection, and will help you to be efficient in your editing.
So, to make the change described above, enter:
docker-compose exec -u 1000 nedi bash
This will place you at a shell in the container as the nedi user, for example:
Change to the mounted configuration volume directory for nedi:
Edit the file with vi:
Make the change(s) and save the file. It is then good practice to check that it is still at the native permissions:
ls -la nedi.conf
You should see something like this:
Leave the shell:
NeDi makes immediate use of the changes you make to its settings file. There's no need to restart the container in this case. You are done!
Using similar procedures, you can access the nagios container, for example:
Using the -u 1000 option puts you in the container as the container's default user (nedi, nagios, etc.). If you leave it out, you will get a root shell, which is not recommended, since you can easily make a change that will leave your container (and perhaps the entire monitoring system) inoperable.
The directories you can access will vary by container.
For the Nagios container, the directories:
are useful for testing, and custom settings, though we caution against adding custom plugins to your system. They will vanish on upgrade!
A better way is to use GDMA (see the GDMA Monitoring documentation).
The global settings can be configured by making changes to the files in /usr/local/groundwork/config.
Configuration files available:
In addition, there are subdirectories that contain more files and templates that you can manage:
The thing to remember is that changing these settings files can affect multiple aspects of the system. Make sure you comment your changes so they are easily reversible, and make regular backups of the system so you can revert the whole thing if you make a catastrophic change.