Chapter 6. Troubleshooting Red Hat CodeReady Containers
The goal of Red Hat CodeReady Containers is to deliver an OpenShift environment for development and testing purposes. Issues occurring during installation or usage of specific OpenShift applications are outside of the scope of CodeReady Containers and should be reported to the relevant project. For example, OpenShift issues are tracked on GitHub.
6.1. Basic troubleshooting
The majority of issues can be resolved by stopping a running CodeReady Containers virtual machine, deleting the virtual machine, and starting a new instance of the virtual machine.
Prerequisites
-
The host machine has been set up using the
crc setup
command. For more information, see Section 3.1, “Setting up CodeReady Containers”. -
The virtual machine has been started using the
crc start
command. For more information, see Section 3.2, “Starting the virtual machine”. - You are using the latest CodeReady Containers release. The generated certificates for the embedded system image bundle expire after 30 days. Using an older system image bundle may result in errors related to expired x509 certificates. For more information, see Section 6.3, “Troubleshooting expired certificates”.
Procedure
To troubleshoot CodeReady Containers, perform the following steps:
Stop the CodeReady Containers virtual machine:
$ crc stop
Delete the CodeReady Containers virtual machine:
$ crc delete
Start the CodeReady Containers virtual machine:
$ crc start
NoteThe cluster takes a minimum of four minutes to start the necessary containers and operators before serving a request.
If your issue has not been resolved by this procedure, perform the following steps:
- Search open issues for the issue that you are encountering.
-
If no existing issue addresses the encountered issue, create an issue and attach the
~/.crc/crc.log
file to the created issue. The~/.crc/crc.log
file contains detailed debugging and troubleshooting information which can help diagnose the problem that you are experiencing.
6.2. Getting shell access to the OpenShift cluster
Direct access to the OpenShift cluster is not needed for regular use and is strongly discouraged. If direct access to the cluster is needed for troubleshooting or debugging purposes, follow this procedure.
Prerequisites
-
oc
access to the cluster is enabled and you are logged in as thekubeadmin
user. For detailed steps, see Section 3.3.2, “Accessing the OpenShift cluster withoc
”.
Procedure
Run
oc get nodes
. The output will be similar to this:$ oc get nodes NAME STATUS ROLES AGE VERSION crc-shdl4-master-0 Ready master,worker 7d7h v1.14.6+7e13ab9a7
-
Run
oc debug nodes/<node>
where<node>
is the name of the node which was printed in the previous step.
6.3. Troubleshooting expired certificates
Prior to Red Hat CodeReady Containers 1.2.0, the system bundle included in each released crc
binary expired 30 days after the release due to certificates embedded in the OpenShift cluster. Using an older crc
binary or system bundle will result in an expired certificates error.
Starting from CodeReady Containers 1.2.0, the embedded certificates can be automatically renewed by crc
. The certificate renewal process is triggered when needed when crc start
is run. When certificates need to be renewed, this can add up to five minutes to the start time of the cluster.
Procedure
With CodeReady Containers releases older than 1.2.0, to resolve expired certificate errors:
-
Download the latest CodeReady Containers release and place the
crc
binary in your$PATH
. Remove the cluster with certificate errors using the
crc delete
command:$ crc delete
WarningThe
crc delete
command will result in the loss of data stored in the CodeReady Containers virtual machine. Save any desired instance information before running this command.Set up the new release:
$ crc setup
Start the new cluster:
$ crc start
6.4. Troubleshooting bundle version mismatch
CodeReady Containers instances are created with bundle information and instance data. When a new CodeReady Containers release is set up, the bundle information and instance data is not updated due to customization in the previous instance data. This will lead to errors when running the crc start
command:
$ crc start ... FATA Bundle 'crc_hyperkit_4.2.8.crcbundle' was requested, but the existing VM is using 'crc_hyperkit_4.2.2.crcbundle'
Procedure
Issue the
crc delete
command before attempting to start the instance:$ crc delete
WarningThe
crc delete
command will result in the loss of data stored in the CodeReady Containers virtual machine. Save any desired instance information before running this command.