- Automation Suite on Linux Release Notes
- Automation Suite on EKS/AKS Release Notes
Automation Suite release notes
2023.10.4
Release date: June 27, 2024
We have introduced more prerequisite checks to optimize the overall experience of installing and configuring Automation Suite and to catch missing requirements earlier. Here are some highlights:
-
Automation Suite now checks if the external objectstore supports POST requests to buckets via pre-signed URLs. Document Understanding requires POST request support to download files from the buckets.
-
We have introduced new prerequisite checks in the
mirror-registry.shandhydrate-registry.shscripts. These checks validate that you have enough free disk space for Docker (under/var/lib/docker) and Podman (under/var/lib/containers) before mirroring or hydrating the registry. Additionally, you are prompted with a message to either continue or abort if you do not have sufficient resources.
4.0.2.
We have improved the Automation Suite support bundle tool capabilities. Now, the tool extracts all the config maps data from your cluster. The new capability allows you to use more valuable data sets for troubleshooting purposes.
We are happy to announce that Automation Suite now supports Harbor and other external registries that require you to create a project before pushing or pulling images from the registry.
We're happy to announce various fixes and improvements that ensure an even more accurate estimate of the hardware requirements for any Automation Suite deployment. The tool now takes into account the additional data disk required by AI Center and Document Understanding. Also, it now takes a single click to share the UiPath Automation Suite Install Sizing Calculator URL along with your currently selected configuration.
If you want to take the UiPath Automation Suite Install Sizing Calculator for a spin, see Capacity planning.
We have made several accessibility improvements as part of our commitment to develop inclusive technology for our users.
- Erratum - added February 27, 2025: We fixed an issue that was preventing document validation tasks, specifically those involving the IsEmpty business rule, from being displayed accurately in Action Center.
- Erratum - added February 27, 2025: In Apps, preview and publish of VB expression apps were not working. We fixed this issue.
- Erratum - added February 27, 2025: In Apps, uploads and downloads of images or documents failed in proxy environments for all apps. We fixed this issue.
-
When generating the support bundle, an incorrect FQDN was used for AKS on Azure Government. We have fixed the issue.
Erratum - added December 18, 2024: An issue causes upgrades to fail when the existing Insights PVC sizes are inadvertently overridden. To address this problem, you must manually change the PVC sizes in ArgoCD UI. For details, see the Troubleshooting section.
This issue is fixed in Automation Suite 2023.10.7.
Erratum - added December 18, 2024: Following an Automation Suite upgrade from version 2023.4 or older, reinstalling or upgrading Insights results in data loss due to an issue with Insights storage class changes. To address the problem, see Troubleshooting.
We fixed the issue in Automation Suite 2023.10.7.
redis-cluster-0 pod to get stuck in the terminating state when you perform a node drain in non-HA Redis scenarios. To address the problem,
you must force delete the pod using the following command:
kubectl -n redis-system delete pod redis-cluster-0 --forcekubectl -n redis-system delete pod redis-cluster-0 --forceWe fixed the issue in Automation Suite 2023.10.7.
redis.<primary-fqdn> and redis-db.<primary-fqdn> during Active/Passive prerequisite checks. This issue is fixed in Automation Suite 2023.10.7.
Erratum - added December 18, 2024: An Insights annotation issue blocks the Automation Suite installer. We fixed the issue in Automation Suite 2023.10.7.
Erratum - added December 18, 2024: The RKE2 log collector runs indefinitely for some system commands. We fixed the issue in Automation Suite 2023.10.7.
exclude= rke2-* is not added to the /etc/yum.conf file on nodes other than the first server. In specific environments, particularly online ones, an attempt to upgrade all
components can cause an unintentional upgrade of the RKE2 service on nodes other than the first server.
exclude=rke2-* to the /etc/yum.conf file on all the nodes of your Automation Suite cluster.
aicenter skill sync and aicenter skill status commands during a side-by-side upgrade.
To fix this issue, you must manually edit the syntax. For details, see the Troubleshooting section.
orchestrator.testautomation section of the cluster_config.json file, the uipathctl binary ignores the connection string and uses the one under the orchestrator section instead. To address the problem, see the Troubleshooting section.
registries.trust.enabled parameter to true in the cluster_config.json file.
rook-ceph OSD PV, the new size does not persist following an Automation Suite upgrade. We fixed the issue in Automation Suite 2023.10.6.
Erratum - added October 17, 2024: When deploying Automation Suite on AWS machines where only IMDSv2 is enabled, installation fails while populating node labels. To address the issue, see the Important note in Optional: Enabling resilience to zonal failures in a multi-node HA-ready production cluster.
System.Net.Http.HttpRequestException: The proxy tunnel request to proxy 'http://<proxyFQDN>:8080/' failed with status code '404'.System.Net.Http.HttpRequestException: The proxy tunnel request to proxy 'http://<proxyFQDN>:8080/' failed with status code '404'.To fix the issue, see the Troubleshooting section.
uipathctl mistakenly processes ports in helm URLs as external registries, excluding the insecure_skip_verify: true parameter from the registries.yaml file. To overcome this issue, you must manually include the parameter in your registries.yaml file. We fixed this issue in Automation Suite 2023.10.6.
rke2-server service is shut down without executing the rke2-killall.sh script. We fixed this issue in Automation Suite 2023.10.6.
- During the installation of Orchestrator certificates for trusted communication.
- When setting up an Identity Service client for Document Manager.
- When configuring an Identity Service client for metering services.
system-upgrade namespace. If the upgrade command fails at any stage (infra, fabric, or service upgrade), take the following steps before
retrying the upgrade:
-
List the existing jobs in the
system-upgradenamespace:kubectl get jobs -n system-upgradekubectl get jobs -n system-upgrade -
Delete the failed jobs:
kubectl -n system-upgrade delete jobs <failed_jobs>kubectl -n system-upgrade delete jobs <failed_jobs>When running the command, replace the<failed_jobs>placeholder with the names of the failed jobs, separated by spaces.
We fixed the issue in Automation Suite 2023.10.5.
Erratum - added August 8, 2024: The bug fix related to document validation tasks in Action Center from Automation Suite 2023.10.3+patch1 was not included in the Automation Suite 2023.10.4 release. We recommend that you wait for Automation Suite 2023.10.5 to get this fix.
Erratum - added August 14, 2024: An issue affects the support bundle generation when using an external OCI registry in an offline environment. Under these conditions, the generated support bundle lacks historical logs and cannot upload to any selected external object storage. To fix this issue, see the Troubleshooting section.
Erratum - added August 14, 2024: Upgrading from Automation Suite 2023.10.x fails due to the Longhorn storage classes still being present. To solve the problem, you must clean up the Longhorn storage classes manually, by running the following command:
for sc in $(kubectl get sc -o json | jq -r '.items[] | select(.provisioner=="driver.longhorn.io") | .metadata.name'); do
kubectl delete sc "${sc}"
done for sc in $(kubectl get sc -o json | jq -r '.items[] | select(.provisioner=="driver.longhorn.io") | .metadata.name'); do
kubectl delete sc "${sc}"
doneWe fixed the issue in Automation Suite 2023.10.5.
ListBucket API prerequisite check fails when using an external objectstore. However, AI Center and Task Mining do not use the ListBucket API permission. We removed the unnecessary check in Automation Suite 2023.10.5.
Erratum - added August 14, 2024: You cannot enable SSO for ArgoCD due to an issue with the Dex image versions. We fixed the issue in Automation Suite 2023.10.5.
snapshot-controller-crds pod to remain in the CrashLoopBackOff state after upgrade. This occurs due to a conflict between the newly installed snapshot-controller and the existing one during the RKE2 upgrade. To fix the issue, see the Troubleshooting section.
Erratum - added August 14, 2024: In certain conditions, the Automation Suite installation process can fail due to telemetry check errors. To fix this issue, you have to manually set the telemetry check as optional and restart the installation. Starting with Automation Suite 2023.10.5, telemetry check errors no longer disrupt the installation process.
Dashboard not found and
Invalid dashboard UID in the request appear in the Grafana
dashboard.
-
Modify the helm parameter for the
fabric-installerapplication. This automatically triggers the sync operation forfabric-installerapplication.kubectl patch application fabric-installer -n argocd --type='json' -p='[{"op": "add", "path": "/spec/source/helm/parameters/-", "value": {"name": "global.rancherMonitoring.grafana.replicas", "value": "1"}}]'kubectl patch application fabric-installer -n argocd --type='json' -p='[{"op": "add", "path": "/spec/source/helm/parameters/-", "value": {"name": "global.rancherMonitoring.grafana.replicas", "value": "1"}}]' - Sync the
rancher-monitoringapplication from ArgoCD UI.
We have fixed this issue in Automation Suite 2023.10.7.
Erratum - added May 21, 2025: For offline Automation Suite installations using the external docker registry, the uipathctl health check command does not work due to a known issue. This issue has been fixed in Automation Suite 2023.10.6.
As a workaround, you can download the uipathctl binary from Automation Suite 2023.10.6 and use it to run the uipathctl health check command.
certificate-injector-webhook in deployment.apps. To address the issue, you must run the following command:
kubectl delete configmap -n uipath-infra certificate-injector-webhook-configmap --ignore-not-foundkubectl delete configmap -n uipath-infra certificate-injector-webhook-configmap --ignore-not-foundWe fixed the issue in Automation Suite 2023.10.9.
du-digitizer-worker deployment from the old Automation Suite version not being fully cleaned up after the upgrade. Starting with Automation Suite
2023.10.0, the deployment was renamed to du-digitizer-worker-deployment.
documentunderstanding application, manually delete the old du-digitizer-worker deployment. After deletion, re-sync the documentunderstanding application.
We fixed the issue in Automation Suite 2023.10.8.
Erratum - added April 15, 2025: An issue prevents the registry cleanup command from effectively clearing the unused Docker images from all registry pods. To address the issue, refer to the Troubleshooting section.
We fixed the issue in Automation Suite 2023.10.9.
Erratum - added April 15, 2025: We have identified an issue impacting the side-by-side and the in-place upgrade of Automation Suite 2022.10 or earlier to Automation Suite 2023.10.0 or later. Due to a faulty migration from MongoDB to SQL Server, the upgrade operation fails if you have Apps enabled and use Kerberos authentication for the SQL Server database.
The recommended solution is to upgrade to Automation Suite 2023.10.9.
node-monitor component, which monitors the node for issues such as checking kube-proxy health or if ip_forward is enabled, and cordons the node if issues arise, is not working for specific nodes such as GPU, task-mining, or as-robot. We have fixed this issue in Automation Suite 2023.10.9.
Erratum - added April 15, 2025: Kerberos tickets fail to automatically renew after cluster restoration, requiring manual sync of the Windows Auth app.
This issue is fixed in Automation Suite 2023.10.9.
uipathctl from correctly setting executeMigration to true during the migration process from Ceph in-cluster storage to external object storage in AI Center.
executeMigration to true in ArgoCD for AI Center and then proceed with a resync, ensuring to prune. This issue is fixed in Automation Suite 2023.10.8.
containerd issue, the deployment of ML skills fails as a token expires before larger image downloads can be completed, especially under
conditions of slower network connectivity.
To address this isssue, see the Troubleshooting section.
Erratum - added February 13, 2025: An issue affects the functionality of email alerts configuration and the setup of multiple receivers after upgrading from Automation Suite 2022.10. To address this issue, see Troubleshooting.
This issue is fixed in Automation Suite 2023.10.8.
We recommend that you regularly check the deprecation timeline for any updates regarding features that will be deprecated and removed.
To find out what has changed on each Automation Suite product, visit the following links.
If the product is greyed out, this new Automation Suite version does not bring any changes to it.
This Automation Suite release bundles the following internal components:
|
Component |
Version |
|---|---|
|
RKE2 |
1.28.7 |
|
ArgoCD |
2.10.9 |
|
logging-operator |
4.6.0 |
| logging-operator-logging |
4.6.0 |
| gatekeeper |
3.16.0 |
|
rook-ceph |
1.9.4 |
| prometheus-pushgateway |
2.12.0 |
|
cert-manager |
1.14.5 |
|
rancher-istio |
102.5.0-up1.19.6 |
| rancher-logging |
103.0.0-up3.17.10 |
| rancher-logging-crd |
103.0.0-up3.17.10 |
| rancher-monitoring-crd |
103.1.0-up45.31.1 |
| rancher-gatekeeper-crd |
103.1.0-up3.13.0 |
| rancher-gatekeeper |
103.1.0-up3.13.0 |
| rancher-monitoring |
103.1.0-up45.31.1 |
|
longhorn |
1.5.5 |
|
longhorn-crd |
1.1.100 |
|
reloader |
1.0.95 |
|
csi-driver-smb |
1.14.0 |
|
velero |
6.2.0 |
|
redis-operator |
7.4.2-12 |
|
redis-cluster |
7.4.2-129 |
The migration tool version you need depends on the standalone products you plan to migrate and the targeted Automation Suite version. For more details, see Migration compatibility matrix.
For instructions on migrating a standalone product to the current version of Automation Suite, see Full migration.
- New prerequisite checks
- Updated Podman version
- Support bundle tool improvement
- Support for external registries that require projects
- Improved UiPath Automation Suite Install Sizing Calculator
- Improved accessibility
- Bug fixes
- Known issues
- Upgrade failure due to overridden Insights PVC sizes
- Data loss when reinstalling or upgrading Insights following Automation Suite upgrade
- Node maintenance issue in non-HA Redis
- Unnecessary Redis DNS resolution validations during Active/Passive checks
- Insights annotation issue blocks installer
- RKE2 log collector runs indefinitely
- Unintended RKE2 service upgrade on additional nodes
- AI Center skills sync failure during side-by-side upgrade
- Test Automation SQL connection string is ignored
- Enabling Connaisseur causes installation or upgrade failures
- OSD PV resize does not persist after upgrade
- Redis failures due to inconsistent password updates
- Redis failures due to inconsistent password updates
- Installation fails while populating node labels
- Resized PVC upgrade issue
- Pods cannot communicate with FQDN in a proxy environment
- Registry configuration corruption during upgrades
- Intermittent not ready reports from agent machines
- AI Center issues with external Orchestrator and Identity Service
- Weak ciphers in TLS 1.2
- Cannot upgrade due to failed jobs in system-upgrade namespace
- Automation Suite 2023.10.3+patch1 issue
- Support bundle log collection failure
- Upgrades fail due to Longhorn being uninstalled
- Unnecessary ListBucket API prerequisite check fails for external objectstore
- Cannot enable SSO for ArgoCD
- snapshot-controller-crds pod in CrashLoopBackOff state after upgrade
- Installation fails due to telemetry checks
- Grafana dashboard error
- uipathctl health check command
- Kubernetes upgrade failure
- Document Understanding deployment issue post-upgrade
- Incomplete Docker registry cleanup process
- Upgrade fails due to MongoDB to SQL Server migration
- Kube-proxy health check not working
- Kerberos tickets do not automatically renew
- AI Center storage migration failure
- ML skill deployment failure due to token expiry
- Email alerts configuration failure post upgrade
- Deprecation timeline
- Bundling details
- Product versions
- Internal third-party component versions
- Migration tool version
