Skip to content

Latest commit

 

History

History
64 lines (49 loc) · 2.8 KB

logging.md

File metadata and controls

64 lines (49 loc) · 2.8 KB
Raw

Logging

Before you begin

The following variables will be used in this guide:

export INSTANCE_NAME=<the name of the instance for which you want to inspect logs>
export NS=<namespace in which your El Carro instance is deployed>
export PATH_TO_EL_CARRO_RELEASE=<the complete path to the downloaded release directory>

You should set these variables in your environment.

Viewing logs via kubectl

You can retrieve El Carro logs from your Kubernetes cluster via the kubectl command, which will display the stdout and stderr streams of the specified container (alert-log-sidecar in this example) by running:

kubectl logs -f $(kubectl get pod -n $NS -l instance=$INSTANCE_NAME -o jsonpath="{.items[0].metadata.name}") -c alert-log-sidecar -n $INSTANCE_NAME

This will display all the stdout/stderr output of the specified container. Aside from the regular agents, we have provided two sidecar containers that tail the listener log and alert log file, named listener-log-sidecar and alert-log-sidecar respectively.

Viewing logs via Cloud Console

You can also retrieve El Carro logs using the Google Cloud Logs Explorer. This feature allows you to view logs of containers for up to 30 days. The Logs explorer allows searching for particular messages, as well as viewing logs for containers that have been shut down or restarted. For more information on Logs Explorer, see the information at the Logs Explorer help page.

Since it may take a few minutes for the database instance to get fully provisioned (largely depending on whether or not the database container has been locally cached or not), the alert log won't show up in Stackdriver until instance provisioning is complete. Once it does (see alert-log-sidecar container's "View Logs" link), there's an option to time lock, jump to now or stream the logs to get the latest updates automatically.

Changing Log Verbosity Levels

The El Carro operator and its agents use the klog logger, which is a fork of glog, and has the ability to dynamically set the verbosity level of the logs. The default level of log verbosity is set to 0, but it can be increased to any positive integer value. Any log messages less than or equal to the current verbosity level will be printed out, for example if the verbosity level is set to 2 then log messages at level 0, 1, or 2 will be printed out, but messages at level 3 or greater will not be printed out.

You're free to change the verbosity level of the operators or config-agent by setting the value of the log levels in the config spec (see the v1alpha1_config_gcp1.yaml file for an example), and then applying it with the command below.

kubectl apply -f $PATH_TO_EL_CARRO_RELEASE/samples/v1alpha1_config_gcp1.yaml -n $NS