Changing the network range of the appliance
In some infrastructures, there can be a collision of IP ranges of the appliance internal network with already existing IP ranges in the surrounding infrastructure. As a result, packet routing may not work correctly. The solution is to regenerate the appliance network interface using a different, non-conflicting network IP range.
At first, we need to find non-conflicting network range of addresses of size /24 which we will use for the new appliance network. Once you have it determined, follow this tutorial.
-
Turn off all the running containers.
systemctl stop iam-cas iam-czechidm iam-web-proxy iam-directory-server iam-czechidm-db
-
Using the command
docker ps
check that no container is running (the list will be empty). -
Delete the internal network using the command.
docker network rm appliance
-
Using
docker network ls
, list all Docker networks and check that the networkappliance
does not exist. -
Create a new network with non-conflicting range,
ADDRESS
is the newly selected non-conflicting range in CIDR format (address/mask
).docker network create --subnet ADDRESS --opt "com.docker.network.bridge.name=appliance" appliance`
-
Using
docker network ls
check the newly created network. -
Delete all original Docker containers because they maintain a link to the original network. Using
docker ps -a
, list all containers. Then, with the commanddocker rm CONTAINER_NAME
, remove them. -
After that, you can gradually start the services. You should check up on their start in the log files.
-
First, start the CzechIdM database
systemctl start iam-czechidm-db
. -
Then, start the directory server
systemctl start iam-directory-server
. -
Start the web proxy and the CAS access manager
systemctl start iam-web-proxy iam-cas
. -
Start CzechIdM
systemctl start iam-czechidm
. Because the container will be newly created, the application will be relinked and the start can take more than 10 minutes. -
Using
docker ps
, check running containers.
-