- Automation Suite on Linux Release Notes
- Automation Suite on EKS/AKS Release Notes
- Automation Suite on OpenShift release notes
2024.10.4
2024.10.4 release notes for Automation Suite on Linux.
Release date: March 26, 2026
Support for new RHEL version
We have expanded our OS support to include RHEL 9.7. For details on the supported RHEL versions, refer to RHEL compatibility matrix.
Release date: June 26, 2025
Support for new RHEL version
We have expanded our OS support to include RHEL 9.6. For details on the supported RHEL versions, see Compatibility Matrix.
Release date: September 24, 2025
Changes to the patching procedure
The procedure for applying patches now includes an additional step that requires you to download and use the corresponding uipathctl patch version before proceeding with the patch.
For more details, refer to Applying a patch.
Deprecation of UiPath Backup based on NFS server
Starting with Automation Suite 2025.10, NFS server-based backups will not be supported if you are using an external objectstore. Your backups will be stored directly in the external objectstore. If you are upgrading to 2025.10, you are encouraged to switch to the external objectstore for backups.
Learn more about Configuring the external objectstore or reach out to support.
Bug fixes
- Erratum - added July 17, 2025: We have fixed an issue where uploading large Document Understanding bundles caused nodes to go out of memory due to parallel loading of large-sized image layers into memory.
- An issue was causing the Task Mining service to fail to initialize in certain Kerberos environments. This occurred due to missing configuration or parameters related to keytab secrets and app settings. We have fixed the issue.
- An issue was causing the addition process of agent nodes to fail in offline environments. We have fixed the issue.
- We have fixed an issue causing frequent restarts of deployments in the
uipathnamespace during offline installations due to inconsistentistio-ingressgateway-certssecrets across namespaces. - We have fixed an issue causing Automation Suite installations to fail due to Certificate Authority (CA) certificates not being recognized. The issue occurred when the
CertificatePoliciessection included policy OID values exceeding 4 bytes. - We have fixed an issue causing the custom changes made to the RKE2 configuration file (
/etc/rancher/rke2/config.yaml) to be overwritten during upgrades. Now, the custom configurations are preserved. - We have fixed an issue where the Process Mining portal failed to load the process apps.
- Log forwarding did not work in proxy setups because the proxy environment variables were not set in the logging pods. We have fixed the issue.
- Licensing SQL connection errors occurred when the Data Source property specified both a named instance and a port. We have fixed the issue.
- We have fixed an issue that caused service disruptions due to automatic secret rotation, making services temporarily inaccessible.
Known issues
Kerberos KDC discovery fails in Active/Active deployments due to truncated DNS SRV responses
Erratum - added September 10, 2026: In Automation Suite Active/Active deployments on Linux (RKE2), Kerberos KDC discovery can fail during installation or upgrade, resulting in Integrated Authentication and LDAP errors. The issue occurs because the Kubernetes DNS Service exposes DNS resolution over UDP/53 only. When a Kerberos SRV lookup, such as _kerberos._tcp or _kerberos._udp, returns a response too large for UDP, the DNS client retries over TCP/53. Since the Kubernetes DNS Service doesn't expose TCP/53, the retry fails.
To address the issue, enable TCP/53 on the Kubernetes DNS Service:
-
Edit the
rke2-corednsHelm chart configuration:kubectl -n kube-system edit helmchartconfig rke2-corednskubectl -n kube-system edit helmchartconfig rke2-coredns -
Under
zoneinspec.valuesContent, adduse_tcp: true:servers: - zones: - zone: . use_tcp: true port: 53servers: - zones: - zone: . use_tcp: true port: 53
We fixed the issue in Automation Suite 2024.10.10.
Monitoring configuration updates do not take effect
Erratum - added July 2, 2026: An issue prevents the monitoring section parameters in cluster_config.json from being applied to the monitoring stack.
This issue is fixed in Automation Suite 2024.10.9.
Prerequisite checks fail when Document Understanding modern projects is enabled and AI Center is disabled
Erratum - added July 2, 2026: An issue causes the prerequisite checks command to incorrectly fail when Document Understanding modern projects is enabled and AI Center is disabled.
To address the issue, refer to Troubleshooting.
This issue is fixed in Automation Suite 2024.10.9.
Custom platform database name not supported when installing Process Mining
Erratum - added March 26, 2026: When Process Mining is enabled, using a custom platform database name causes the OMS backfill job to fail.
As a workaround, use AutomationSuite_Platform as the platform database name when installing Process Mining.
Redis telemetry job pods failing in airgapped environments
Erratum - added March 26, 2026: Redis telemetry is enabled by default but the required image is not included in the offline bundle, causing ImagePullBackOff errors in airgapped environments. You can either ignore the errors or delete the Redis telemetry cronjob.
We fixed the issue in Automation Suite 2024.10.8.
Prerequisite check incorrectly validates monitoring and logging components
Erratum - added March 26, 2026: The prerequisite check incorrectly validates monitoring and logging components. Since these components are installed after the prerequisite phase, this validation is not necessary.
We fixed the issue in Automation Suite 2024.10.8.
Insights real-time monitoring not working on the secondary cluster in Active/Active deployments
Erratum - added March 26, 2026: In Active/Active deployments with Insights real-time monitoring enabled, Orchestrator and platform services on the secondary cluster do not push real-time monitoring data due to configuration issues with call redirection from the secondary cluster to the primary cluster.
We fixed the issue in Automation Suite 2024.10.8.
FIPS 140-2 support limitations
Erratum - added January 22, 2026: Insights is not supported on Automation Suite deployments that run on FIPS 140-2-enabled machines. To remain compliant with FIPS 140-2 requirements, you must disable Insights.
For details, refer to Security and compliance.
AI Center in-cluster proxy connectivity issue
Erratum - added December 19, 2025: An issue prevents AI Center pods from connecting to the in-cluster object store in proxy-enabled environments. This occurs due to unsupported CIDR notations and FQDN regex patterns in the no_proxy configuration.
To address the issue, refer to the Troubleshooting section.
We fixed the issue in Automation Suite 2024.10.7.
Prerequisite checks fail when Process Mining is enabled
Erratum - added December 19, 2025: An issue causes the prerequisite checks to fail when Process Mining is enabled. The issue occurs due to the checks still requiring cert-manager, although cert-manager is no longer needed for Process Mining starting with Automation Suite 2024.10.3.
The failure message is a false positive and can be safely ignored.
We fixed the issue in Automation Suite 2024.10.7.
Pre-upgrade command fails with proxy and external objectstore
Erratum - added December 19, 2025: An issue prevents the uipathctl cluster pre-upgrade command from completing successfully in environments configured with a proxy and external objectstore. The issue occurs due to an error during the Insights volume migration process.
We fixed the issue in Automation Suite 2024.10.7.
Upgrade failure during posthook import in chained upgrades
Erratum - added December 19, 2025: An issue causes the upgrade to fail during the execution of the uipathctl rke2 upgrade command in chained upgrade scenarios. This issue occurs due to the posthook secret and configmap import operation not being idempotent, which leads to conflicts with existing Kubernetes objects.
To address the issue, refer to the Troubleshooting section.
We fixed the issue in Automation Suite 2024.10.7.
Kerberos keytab rotation does not trigger token regeneration
Erratum - added September 24, 2025: An issue causes Kerberos keytab rotation to not immediately regenerate authentication tokens, which may lead to temporary connectivity disruptions between services and the database.
To address the issue, you must manually trigger the Kerberos ticket renewal cronjob by running the following command:
kubectl delete job tgt-rotate-manual -n uipath --ignore-not-found ; kubectl create job tgt-rotate-manual --from=cronjob/kerberos-tgt-update -n uipath
kubectl delete job tgt-rotate-manual -n uipath --ignore-not-found ; kubectl create job tgt-rotate-manual --from=cronjob/kerberos-tgt-update -n uipath
We fixed the issue in Automation Suite 2024.10.5.
Registry certificates not fully updated in offline scenarios
Erratum - added September 24, 2025: An issue causes the uipathctl config tls-certificates command to update only the internal certificate, while missing the certificate required by Argo CD to trust the in-cluster registry in offline scenarios.
To address the issue, you must run the following command to explicitly update the ArgoCD registry certificate in internal–external registry scenarios:
./bin/uipathctl config argocd ca-certificates update --cacert [PATH]
./bin/uipathctl config argocd ca-certificates update --cacert [PATH]
We fixed the issue in Automation Suite 2024.10.5.
GPU enablement fails with external registries containing project names
Erratum - added September 24, 2025: An issue prevents GPU enablement after adding a GPU node when using external registries with project names (such as Harbor). The required pods do not start and display an ImagePullBackoff error.
If you are using an external registry with a project name, update the NVIDIA device plugin daemonset with the following command:
# Replace REGISTRY_WITH_PROJECT_NAME with the correct value (Eg. my.registry.io:443/myproject)
# Replcae TAG with the correct value. You can get this from the <installer_directory>/versions/docker-images.json file (Eg. v0.17.1)
kubectl set image daemonset/nvidia-device-plugin-daemonset \
-n kube-system \
nvidia-device-plugin-ctr=<REGISTRY_WITH_PROJECT_NAME>/k8s-device-plugin:<TAG>
# Replace REGISTRY_WITH_PROJECT_NAME with the correct value (Eg. my.registry.io:443/myproject)
# Replcae TAG with the correct value. You can get this from the <installer_directory>/versions/docker-images.json file (Eg. v0.17.1)
kubectl set image daemonset/nvidia-device-plugin-daemonset \
-n kube-system \
nvidia-device-plugin-ctr=<REGISTRY_WITH_PROJECT_NAME>/k8s-device-plugin:<TAG>
We fixed the issue in Automation Suite 2024.10.5.
Orchestrator does not start when using user-assigned managed identity
Erratum - added September 24, 2025: An issue prevents the storage.useClientID parameter from being set when you select a user-assigned managed identity for object storage. As a result, Orchestrator cannot start in environments where access is restricted to user-assigned managed identities.
To address the issue, you must manually set the parameter in ArgoCD, as follows:
- In the ArgoCD Orchestrator application, go to Details > Parameters.
- In the values text box, set
storage.isExternal : trueand save.
We fixed the issue in Automation Suite 2024.10.5.
Temporary registry installation timeout
Erratum - added September 24, 2025: An issue causes the temporary registry installation to fail with a timeout error when connecting to registry-1.docker.io. To address the issue, refer to the Troubleshooting section.
We fixed the issue in Automation Suite 2024.10.5.
Thanos compactor stops on corrupted blocks
Erratum - added September 24, 2025: An issue causes the Thanos compactor to stop compacting metrics when it encounters corrupted blocks in the object store. This prevents compaction and leads to increased storage usage.
To address the issue, refer to the Troubleshooting section.
We fixed the issue in Automation Suite 2024.10.5.
Upgrade fails during rook-ceph app deletion
Erratum - added September 24, 2025: An issue causes the upgrade to fail at the pre-upgrade stage and get stuck while deleting the rook-ceph ArgoCD application. This issue occurs when upgrading from Automation Suite 2023.10.10 to version 2024.10.4.
To address the issue, refer to the Troubleshooting section.
We fixed the issue in Automation Suite 2024.10.5.
Garbage collection fails during registry cleanup
Erratum - added September 24, 2025: An issue prevents garbage collection from executing correctly during in-cluster registry cleanup. This may result in unused images not being cleaned up as expected.
To address the issue, refer to the Troubleshooting section.
We fixed the issue in Automation Suite 2024.10.5.
Deprecation timeline
We recommend that you regularly check the deprecation timeline for any updates regarding features that will be deprecated and removed.
Bundling details
Product versions
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.
Internal third-party component versions
This Automation Suite release bundles the following internal components:
| Component | Version |
|---|---|
| RKE2 | 1.32.3+rke2r1 |
| ArgoCD | v3.0.0 |
| gatekeeper | 3.19.1 |
| rook | 1.17.1 |
| ceph | 19.2.2 |
| prometheus-pushgateway | v3.1.3 |
| cert-manager | v1.17.2 |
| kube-logging/logging-operator | 5.3.0 |
| kube-logging/config-reloader | 5.3.0 |
| istio | 1.25.2 |
| velero | 1.16.0 |
| reloader | v2.1.3 |
| Prometheus | v3.3.0 |
| Grafana | 11.6.1 |
| redis-operator | v7.22.0-7 |
| redis-cluster | v7.22.0-28.focal |
| oauth2-proxy | v7.9.0 |
Migration tool version
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.
- Support for new RHEL version
- Support for new RHEL version
- Changes to the patching procedure
- Deprecation of UiPath Backup based on NFS server
- Bug fixes
- Known issues
- Kerberos KDC discovery fails in Active/Active deployments due to truncated DNS SRV responses
- Monitoring configuration updates do not take effect
- Prerequisite checks fail when Document Understanding modern projects is enabled and AI Center is disabled
- Custom platform database name not supported when installing Process Mining
- Redis telemetry job pods failing in airgapped environments
- Prerequisite check incorrectly validates monitoring and logging components
- Insights real-time monitoring not working on the secondary cluster in Active/Active deployments
- FIPS 140-2 support limitations
- AI Center in-cluster proxy connectivity issue
- Prerequisite checks fail when Process Mining is enabled
- Pre-upgrade command fails with proxy and external objectstore
- Upgrade failure during posthook import in chained upgrades
- Kerberos keytab rotation does not trigger token regeneration
- Registry certificates not fully updated in offline scenarios
- GPU enablement fails with external registries containing project names
- Orchestrator does not start when using user-assigned managed identity
- Temporary registry installation timeout
- Thanos compactor stops on corrupted blocks
- Upgrade fails during rook-ceph app deletion
- Garbage collection fails during registry cleanup
- Deprecation timeline
- Bundling details
- Product versions
- Internal third-party component versions
- Migration tool version