Troubleshooting containers
Refer to the following topics to troubleshoot problems that might occur during or after the installation.
- TrueSight Orchestration peer is not available on the correct port
- No clear error messages appear if adapter requests fail
- Deployment script fails to connect to the Docker registry
- Installation is complete, container is running, but repository is not accessible
- Installation is complete, and container is running, but CDP is not accessible
- TrueSight Orchestration components are removed, volumes are deleted, but containers are still running
TrueSight Orchestration peer is not available on the correct port
When you install TrueSight Orchestration Platform in containers, depending on the Docker version that you are using, Grid Manager will not be available on the port provided during installation. By default, a configuration distribution peer (CDP) is installed on port 38080. However, recent versions of Docker use a Traefik version, which changes the port number to a default 30000. In such a case, even if you provide any port during installation, the port number may be ignored.
If your Grid Manager is not accessible using port number 38080 or any port number that you provided, try and access the Grid Manager using the 30000 port.
Example: https://server.bmc.com:30000/baocdp/gm
For more information, see the Knowledge Article on BMC Communities.
No clear error messages appear if adapter requests fail
If the adapter request contains incorrect inputs and the request fails, the error message in the output is not clear. For example, while using a File adapter, if the file is not found, instead of a message that describes that file is not found, an incomplete message – Unable to carry out request : Message[summary=, detail=] is displayed.
This issue occurs when you redeploy the Configuration Distribution Peer on a container. When you redeploy the CDP, the messages folder that contains all the message strings (located AO_HOME\CDP\messages) is not getting created. To resolve this issue, perform the following steps:
- Open a terminal and SSH into the host where your CDP is installed.
Go to the container where the CDP is installed and run the following command.
# Yum install unzipGo to the /opt/bmc/BAO/CDP/tomcat/webapps/baocdp/WEB-INF/lib location and unzip the resources-8.2.00.jar file.
# -j resources-8.2.00.jar -d ../../../../../messages/ "*.xml"The messages folder is now created in the CDP directory.
Restart the CDP service.
Command to start the CDP service# /opt/bmc/BAO/CDP/bin/server.sh start
Deployment script fails to connect to the Docker registry
During installation, when you provide the registry host and port to pull images from the registry, sometimes, an error might occur.
This issue occurs if you are using an untrusted registry and the registry details are not added to the daemon.json file on a host where you are installing the component. To resolve this issue, perform these steps:
- Go to the /etc/docker/ and add a new daemon.json file.
Add the registry name in the file using the following format:
{
"insecure-registries:
[
"registry_hostname:port"
]
}If using multiple registries, specify a comma-separated list of registries as shown in the following format:
{ "insecure-registries":["registry_hostname:port","registry_hostname:port","registry_hostname:port"] }Save changes to the file and restart Docker Service.
# service docker restart
Installation is complete, container is running, but repository is not accessible
If your installation is complete, and the container is running, but the repository is still not accessible, verify whether the repository service is started. If it is not, start the repository service.
Go to the docker container where the repository is installed using this command:
Command to enter container# docker exec -it <container ID> bashYou are now inside the container.
Go to the folder where the repository is installed and start the service using this command:
Command to start the repository service# /opt/bmc/BAO/REPO/bin/server.sh startThe repository service is started.
- Go the URL and log on to the repository.
For instructions, see Accessing-and-navigating-the-Repository-Manager.
Installation is complete, and container is running, but CDP is not accessible
If your installation is complete and the CDP is still not started, verify whether the CDP service is started. If it is not, start the CDP service.
Go to the docker container where the CDP is installed using this command:
Command to enter container# docker exec -it <container ID> bashYou are now inside the container.
Go to the folder where the repository is installed and start the service using this command:
Command to start the CDP service# /opt/bmc/BAO/CDP/bin/server.sh startThe CDP service is started.
- Go the URL and log on to the repository.
For instructions, see Accessing-and-navigating-the-Grid-Manager.
TrueSight Orchestration components are removed, volumes are deleted, but containers are still running
If you have chosen to manually remove the TrueSight Orchestration components from a host, TrueSight Orchestration components and Docker volumes are removed but sometimes, containers may not be removed from your environment. This may cause issues when you want to install TrueSight Orchestration in containers again.
This may occur on an RHEL operating system due to a Docker known issue (Containers not removed after docker stack rm), where data from the local volumes is not cleared. As a workaround, you must remove the containers manually. For more information, see docker rm in Docker documentation.
BMC also recommends that you use the maintenance tool (provided with the TrueSight Orchestration installation files) to remove or maintain your container environment. For instructions, see Running-the-maintenance-tool-for-TrueSight-Orchestration-in-containers.