There are a couple of issues related to changing the logging level for the Boot Orchestration Service (BOS).
After changing the BOS logging level, the BOS API server pod logs show a message reflecting this change. However, the server does not fully pick up the change until the Kubernetes deployment of the API server is restarted.
NOTE: The BOS Operators do not have this issue – they pick up the log level change without having to be restarted.
This issue exists in BOS in CSM 1.3 through CSM 1.6. It is fixed in CSM 1.7.
After changing the BOS logging level, the BOS Kubernetes pods show a message reflecting this change. There is an issue in which the new log level is displayed as an integer rather than a string.
The following is an example of what the message should look like:
Logging level changed from INFO to DEBUG
When this issue is present, the message looks like the following:
Logging level changed from INFO to 10
For the BOS API server, this bug exists in CSM 1.6.0 and CSM 1.6.1; it is fixed in CSM 1.6.2. For the BOS operators, this bug exists in CSM 1.6; it is fixed in CSM 1.7.