When observing the failures on EKF platform, it's necessary to collect logs. These logs can then be submitted along with the support ticket. Below are a few different approaches for collecting the logs based on need.
1. Full platform logs. These logs are useful when seeing any failure on EKF platform. Please note that this refers to 1.4 documentation, but this tool can be used to gather log for 1.3 as well. For 1.3 cluster please create a new 1.4 rok-tool pod in your existing 1.3 clusters to run through the following guide.
2. Task logs. Kubeflow UI lists the tasks related to the snapshot process. In case of failures in task it might be useful to look at the full task log to understand the actual reason for failure for those tasks.
3. Browser Network logs. In certain cases where an error pops up on the browser itself, it is useful to collect network logs from the browser to debug the reason for failure.
If you find an error in this article please leave a comment and one of our support staff will perform a review.