Configuring Cluster Snapshot Fails For An Offline Automation Suite Installation

News Knowledge Base
1

Configuring Cluster Snapshot Fails For a Offline Automation Suite Installation.

Error Description: Start Automation Suite v22.10, use Velero for taking the cluster backup and on running the below sample command to configure the cluster snapshot:

  • ./configureUiPathAS.sh snapshot config -s "*/45 * * * *" -r "48h" -t nfs -e -l "/" --prefix "schedule"

However, while running this command it abruptly stops at the "helm registry login" step as shown below,


The command stops here and it is observed that the Velero installation and configuration does not happen.


Error Message:
Snapshot Configure Script Failure Error:
e1.jpg


Manual Helm Login Error:
e2.jpg


Resolution:
This issues occurs as the helm login to the internally hosted offline registry fails.
  1. To test the login, navigate to the bin folder within the installation folder (/opt/UiPathAutomationSuite) and attempt a manual helm login like below:
11.jpg
  1. As seen above the login is failing due to a certificate issue, this is happening because the certificate authority, which issued the server certificate for Automation Suite, is unknown/not trusted by our underlying hosting server. During the Automation Suite Installation the installer places the root certificate (default self-signed created at installation) at this path.
12.jpg

  1. This however can also happen with Users using their AD issued certificates, the reason in that although it is an internal AD issued trusted certificate yet the root authority is not trusted by the server hosting Automation Suite. To bypass this, we can extract/copy the root authority certificate and copy it under /etc/pki/ca-trust/source/anchors and then run update-ca-trust.
  2. Once the root authority is trusted the helm login starts working:
13.jpg
  1. After this the snapshot configure script runs as expected and completes the required Velero installation as well.