Using the Automation Suite support bundle

Overview

The Automation Suite support bundle collects application-level logs of UiPath® services and and relevant objects required to evaluate and fix the problem. The Automation Suite support bundle also contains historical logs you can use to analyze patterns when troubleshooting a problem. Historic logs are particularly helpful when the potential root cause of an issue is not evident just by looking at the state of the cluster.

The Automation Suite support bundle is also needed when asking for guidance or a fix from the UiPath® Support team as they need it for analysis purposes.

Note:

The Automation Suite support bundle does not collect or store any PII or confidential data of you, your user, or your automation. UiPath® applications not not leak or log any confidential information in their records. At the platform level, the log collector is designed to mask any such critical information.

Generating the support bundle

To generate the support bundle, run the following command from the installer directory:

./bin/uipathctl health bundle cluster_config.json --versions versions/helm-charts.json

This tool requires some time to gather and store all the logs in a temporary location on your local or management machine.

The following table lists the optional flags that you can use for the support bundle tool:

Flags	Description
--output-dir string	Specify the directory to store the support bundle.
--skip-diagnose	Boolean flag to skip the diagnostics report. By default, the bundle contains the diagnostics reports. You can use this flag to exclude the diagnostics report from the support bundle.
--namespace string	Specify the namespace where UiPath Automation Suite is deployed.
--include string	Specify additional components (comma separated) to include in the bundle (such as --include=historical-logs).
--exclude string	Specify components (comma separated) to exclude from the bundle (such as--exclude=rke2-bundle).
--from YYYY-MM-DD	The start date to collect the historical logs in YYYY-MM-DD format.
--days int	The number of days from the start date for which to collect the historical logs.
--limit-bytes int	Limits the size of the logs of running pods by providing the integer value in bytes.

By default, if the --from flag is not specified, then the historical logs will be collected for the past 2 days starting from today, irrespective of the --days specified. If the --from is specified and --days is not specified, then the historical logs will be collected for 2 days starting from the specified start date.

The RKE2 bundle is included by default in the support bundle. To exclude it, use theflag --exclude=rke2-bundle.

The historical logs are not included by default in the support bundle. To include them, use the flag --include=historical-logs.

Automation Suite Support Bundle Structure

The .tar.gz archive contains following files and folders:

File/folder	Description
logs	Contains the historical logs collected from ceph or external object store.
current-logs	Contains the live logs (such as logs for current and previous instance of pods).
events.json	Contains the event descriptions from all the namespaces.
alerts	Contains the active prometheus alerts in the cluster.
server0.tar.gz, server1.tar.gz, ...	Contains the RKE2 bundle for every node in the cluster.
namespace/resource/object.yaml	Contains the object descriptions for the server-preferred/custom namespace and cluster-scoped resources.

The RKE2 bundle structure is as follows:

Directory	Description
systemlogs	Contains the logs related to system events, security audits, and general system messages.
systeminfo	Contains the information about the running processes, disk usage, cpu utilisation and so on.
etcd	Contains the information related to the data storage of the etcd key-value store.
journald	Contains the logs managed by systemd-journald.
networking	Contains the network configuration details such as iptables, routing and so on.
rke2	Contains the kubelet logs, agent and server certificates, info on crictl and the resources running on the cluster and the pod logs.