Upgrading AI Center

To upgrade AI Center from one minor version to another (for example, from 2021.10.1 to 2021.10.2), use the procedures from this page.

Note: Make sure that there are no new deployments while upgrading.

Important:

If you upgrade from AI Center 2021.10.3 or 2021.10.4 to a newer version, a downgrade from Ceph 16.2.6 to 15.2.9 is required to prevent data corruption.

For more information, check the Downgrading Ceph from 16.2.6 to 15.2.9 page.

Deployment mode	Upgrade instructions
Online single-node evaluation	Online single-node evaluation mode
Offline single-node evaluation	Offline single-node evaluation mode
Online multi-node HA-ready production	Online multi-node HA-ready production mode
Offline multi-node HA-ready production	Offline multi-node HA-ready production mode

Online Single-node Evaluation Mode

Preparation

1. Make sure that there is enough disk space on the node.
2. Download and unzip the new installer (installer.zip) on the server.
3. Make the original cluster_config.json file available on the server.
4. Generate the new cluster_config.json file as follows:

• If you have the old cluster_config.json , use the following command to generate the configuration file from the cluster:

cd /path/to/new-installer ./configureUiPathAS.sh config get -i /path/to/old/cluster_config.json -o /path/to/store/generated/cluster_config.json

• If you do not have the old cluster_config.json file, run the following command:

cd /path/to/new-installer ./configureUiPathAS.sh config get -o /path/to/store/generated/cluster_config.json

Note: For details on how to configure the cluster_config.json parameters, see Using the configuration file.

Execution

Maintenance and backup

1. Make sure you have enabled the backup on the cluster. For details, see Backing up and restoring the cluster.
2. Connect to the server node via SSH.
3. Verify that all desired volumes have backups in the cluster by running the following command:

   /path/to/new-installer/configureUiPathAS.sh verify-volumes-backup

   Note: The backup might take some time, so wait for approximately 15-20 minutes, and then verify the volumes backup again.
4. Put the cluster in maintenance mode as follows:
5. Execute the following command:

   /path/to/new-installer/configureUiPathAS.sh enable-maintenance-mode

6. Verify that the cluster is in maintenance mode by running the following command:

   /path/to/new-installer/configureUiPathAS.sh is-maintenance-enabled

7. Make an SQL database backup.

Upgrade infrastructure and services on servers

1. Connect to the server via SSH.
2. Become root by running sudo su -.
3. Remove redis section in patch json file.
   Because of a known issue in upgrade we manually need to edit the /opt/UiPathAutomationSuite/Installer/Upgrade/paths/patch.json file. This is fixed in 2022.4 version.
   To do so, open file using a text editor and remove the lines 22 to 25 and the ending , in line 21:

   vi /opt/UiPathAutomationSuite/Installer/Upgrade/paths/patch.json
   remove the lines: 
   {
       "upgradeScriptName": "redis_remove.sh",
       "upgradeScriptArguments": ""
   }
   remove comma on above line.

   After the changes the files should look as below:

   {
       "minSourceVersion": "<version_number>",
       "maxSourceVersion": "<version_number>",
       "minTargetVersion": "<version_number>",
       "maxTargetVersion": "<version_number>",
       "pre_infra_steps": [
           {
               "upgradeScriptName": "longhorn-single-to-multiple-replicas.sh",
               "upgradeScriptArguments": ""
           }
       ],
       "post_infra_steps": [],
       "pre_fabric_steps": [
           {
               "upgradeScriptName": "delete-immutable-jobs.sh",
               "upgradeScriptArguments": ""
           },
           {
               "upgradeScriptName": "rook-ceph-pre-1.7.9.sh",
               "upgradeScriptArguments": ""
           }
       ],
       "post_fabric_steps": [
           {
               "upgradeScriptName": "rook-ceph-post-1.7.9.sh",
               "upgradeScriptArguments": ""
           }
       ],
       "pre_services_steps": [],
       "post_services_steps": []
   }

4. Upgrade the infrastructure and servers by running the following command:

   /path/to/new-installer/install-uipath.sh --upgrade --install-standalone-aicenter -k -i /path/to/cluster_config.json --accept-license-agreement -o /path/to/output.json

5. Upgrade fabric by running the following command:

   /path/to/new-installer/install-uipath.sh --upgrade --install-standalone-aicenter -f -i /path/to/cluster_config.json --accept-license-agreement -o /path/to/output.json

6. Run the software installer by using the command below.

   First you need to connect to identity server and retrieve an installation token. Then copy this token and paste it in placeholder from the `cluster_config.json file.

   Once the token is there, you can run the AI Center install.

   /path/to/new-installer/install-uipath.sh --upgrade --install-standalone-aicenter -s -i /path/to/cluster_config.json --accept-license-agreement -o /path/to/output.json

   Note: This command disables the maintenance mode that you enabled before the upgrade because all services are required to be up during the upgrade. The command also creates a backup of the cluster state and pauses all other scheduled backups.
7. After the successful upgrade and verification, resume the backup scheduling on the node by running the following command:

   /path/to/new-installer/configureUiPathAS.sh resume-scheduled-backups

Rollback on Error

Preparation

1. Create a separate folder to store older bundles, and perform the following operations inside that folder.
2. Download and unzip the installer of the older version (installer.zip) on the node.
   Note: Give proper permissions to the folder by running sudo chmod 755 -R <installer-folder>.
3. Create a restore.json file and copy it to all the nodes. For details, see Backing up and restoring the cluster.
4. Verify that the etcd backup data is present on the primary server at the following location: /mnt/backup/backup/<etcdBackupPath>/<node-name>/snapshots.
   • etcdBackupPath - this is the same as the one specified in backup.json while enabling the backup's node-name;
   • node-name - the hostname of the primary server VM.

Cluster cleanup

1. Copy and run the dedicated script to uninstall everything from that node. Do this for all the nodes. For details, see the How to uninstall the cluster section in the Troubleshooting page.
2. Restore all UiPath databases to the backup created before the upgrade.

Restore infra on server nodes

1. Connect to the server via SSH.
2. Restore infra by running the following command:

   /path/to/older-installer/install-uipath.sh -i /path/to/restore.json -o /path/to/output.json -r --accept-license-agreement --install-type online

Restore volumes data

1. Connect to the server via SSH.
2. Go to the new installer folder.
   Note: The previous infra restore commands were executed using the old installer, and the following commands are executed using the new installer bundle.
3. Disable the maintenance mode on the cluster by running the following command:

   /path/to/new-installer/configureUiPathAS.sh disable-maintenance-mode

4. Verify that maintenance mode is disabled by running the following command:

   /path/to/new-installer/configureUiPathAS.sh is-maintenance-enabled

5. Copy the restore.json file that was used in the infra restore stage to the new installer bundle folder.
6. Restore volumes by running the following command from the newer installer bundle:

   /path/to/new-installer/install-uipath.sh -i /path/to/new-installer/restore.json -o /path/to/new-installer/output.json -r --volume-restore --accept-license-agreement --install-type online

7. Once the restore is completed, verify if everything is restored and working properly.
8. During the upgrade, scheduled backups were disabled on the primary node. To enable them again, run the following command:

   /path/to/new-installer/configureUiPathAS.sh resume-scheduled-backups

Offline Single-node Evaluation Mode

Preparation

1. Make sure that there is enough disk space on the node.
2. Download the full offline bundle (sf.tar.gz) on the chosen server.
3. Download and unzip the new installer (installer.zip) on the primary server.
   Note: Give proper permissions to the folder by running sudo chmod 755 -R <installer-folder>.
4. Make the original cluster_config.json file available on the server.
5. Generate the new cluster_config.json file as follows:
   • If you have the old cluster_config.json , use the following command to generate the configuration file from the cluster.

cd /path/to/new-installer ./configureUiPathAS.sh config get -i /path/to/old/cluster_config.json -o /path/to/store/generated/cluster_config.json

• If you do not have the old configuration file, run the following command:

cd /path/to/new-installer ./configureUiPathAS.sh config get -o /path/to/store/generated/cluster_config.json

Note: See Using the configuration file to fill up the remaining parameters.

Execution

Maintenance and backup

1. Make sure you have enabled the backup on the cluster.
2. Connect to server node via SSH.
3. Verify that all volumes have backups in the cluster by running the following command:

   /path/to/new-installer/configureUiPathAS.sh verify-volumes-backup

   Note: The backup might take some time, so wait for approximately 15-20 minutes, and then verify the volumes backup again.
4. Put the cluster in maintenance mode as follows:
5. Execute the following command:

   /path/to/new-installer/configureUiPathAS.sh enable-maintenance-mode

6. Verify that the cluster is in maintenance mode by running the following command:

   /path/to/new-installer/configureUiPathAS.sh is-maintenance-enabled

7. Create the SQL database backup.

Upgrade infrastructure and services on servers

1. Connect to server via SSH.
2. Become root by running sudo su -.
3. Remove redis section in patch json file.
   Because of a known issue in upgrade we manually need to edit the /opt/UiPathAutomationSuite/Installer/Upgrade/paths/patch.json file. This is fixed in 2022.4 version.
   To do so, open file using a text editor and remove the lines 22 to 25 and the ending , in line 21:

   vi /opt/UiPathAutomationSuite/Installer/Upgrade/paths/patch.json
   remove the lines: 
   {
       "upgradeScriptName": "redis_remove.sh",
       "upgradeScriptArguments": ""
   }
   remove comma on above line.

   After the changes the files should look as below:

   {
       "minSourceVersion": "<version_number>",
       "maxSourceVersion": "<version_number>",
       "minTargetVersion": "<version_number>",
       "maxTargetVersion": "<version_number>",
       "pre_infra_steps": [
           {
               "upgradeScriptName": "longhorn-single-to-multiple-replicas.sh",
               "upgradeScriptArguments": ""
           }
       ],
       "post_infra_steps": [],
       "pre_fabric_steps": [
           {
               "upgradeScriptName": "delete-immutable-jobs.sh",
               "upgradeScriptArguments": ""
           },
           {
               "upgradeScriptName": "rook-ceph-pre-1.7.9.sh",
               "upgradeScriptArguments": ""
           }
       ],
       "post_fabric_steps": [
           {
               "upgradeScriptName": "rook-ceph-post-1.7.9.sh",
               "upgradeScriptArguments": ""
           }
       ],
       "pre_services_steps": [],
       "post_services_steps": []
   }

4. Upgrade the infrastructure by running the following command:

   /path/to/new-installer/install-uipath.sh --upgrade --install-standalone-aicenter -k -i /path/to/cluster_config.json --offline-bundle "/path/to/sf.tar.gz" --offline-tmp-folder /uipath/tmp --install-offline-prereqs --accept-license-agreement -o /path/to/output.json

5. Upgrade fabric by running the following command:

   /path/to/new-installer/install-uipath.sh --upgrade --install-standalone-aicenter -f -i /path/to/cluster_config.json --offline-bundle "/path/to/sf.tar.gz" --offline-tmp-folder /uipath/tmp --install-offline-prereqs --accept-license-agreement -o /path/to/output.json

6. Run the software installer by using the command below.

   First you need to connect to identity server and retrieve an installation token. Then copy this token and paste it in placeholder from the `cluster_config.json file.

   Once the token is there, you can run the AI Center install.

   /path/to/new-installer/install-uipath.sh --upgrade --install-standalone-aicenter -s -i /path/to/cluster_config.json --offline-bundle "/path/to/sf.tar.gz" --offline-tmp-folder /uipath/tmp --install-offline-prereqs --accept-license-agreement -o /path/to/output.json

   Note: This command disables the maintenance mode that you enabled before the upgrade because all services are required to be up during the upgrade. Note that this command also creates a backup of the cluster state and pauses all other scheduled backups.

7. After the successful upgrade and verification, resume the backup scheduling on the node by running the following command:

   /path/to/new-installer/configureUiPathAS.sh resume-scheduled-backups

Rollback on Error

Preparation

1. Create a separate folder to store old bundles, and perform the following operations inside that folder.
2. Download the infra-only offline bundle (sf-infra-bundle.tar.gz) corresponding to the old version on all the nodes.
3. Download and unzip the installer of the old version (installer.zip) on all the nodes.
   Note: Give proper permissions to the folder by running sudo chmod 755 -R <installer-folder>.
4. Create a restore.json file and copy it to the node. For details, see Backing up and restoring the cluster.
5. Verify that the etcd backup data is present on the primary server at the following location: /mnt/backup/backup/<etcdBackupPath>/<node-name>/snapshots.
   • etcdBackupPath - this is the same as the one specified in backup.json while enabling the backup's node-name;
   • node-name - the hostname of the primary server VM.

Cluster cleanup

1. Copy and run the dedicated script to uninstall everything from that node. For details, see the How to uninstall the cluster section in the Troubleshooting page.
2. Restore all UiPath databases to the older backup created before the upgrade.

Restore infra on server nodes

1. Connect to the server node.
2. Restore infra by running the following command:

   /path/to/older-installer/install-uipath.sh -i /path/to/restore.json -o /path/to/output.json -r --offline-bundle "/path/to/older-version/sf-infra-bundle.tar.gz" --offline-tmp-folder /uipath --install-offline-prereqs --accept-license-agreement --install-type offline

Restore volumes data

1. Connect to the server via SSH.
2. Go to the new installer folder.
   Note: The previous infra restore commands were executed using the old installer, and the following commands are executed using the new installer bundle.
3. Disable the maintenance mode on the cluster by running the following command:

   /path/to/new-installer/configureUiPathAS.sh disable-maintenance-mode

4. Verify that maintenance mode is disabled by executing the following command:

   /path/to/new-installer/configureUiPathAS.sh is-maintenance-enabled

5. Copy the restore.json file that was used in the infra restore stage to the new installer bundle folder.
6. Restore volumes from the new installer bundle by executing the following command:

   /path/to/new-installer/install-uipath.sh -i /path/to/new-installer/restore.json -o /path/to/new-installer/output.json -r --volume-restore --accept-license-agreement --install-type offline

7. Once the restore is complete, verify if everything is restored and working properly.
8. During the upgrade, scheduled backups were disabled on this node. To enable them again, run the following command:

   /path/to/new-installer/configureUiPathAS.sh resume-scheduled-backups

Online Multi-node HA-ready Production Mode

Preparation

1. Identify any server (not agent) that meets the disk requirements for an online installation. Referenced as primary server throughout this document. If you are using a self-signed certificate, run the following command:

   <h3>Please replace /path/to/cert with path to location where you want to store certificates.
   sudo ./configureUiPathAS.sh tls-cert get --outpath /path/to/cert
   ### Now copy the ca.crt file generated in above location to trust store location
   sudo cp --remove-destination /part/to/cert/ca.crt /etc/pki/ca-trust/source/anchors/
   ### Update the trust store
   sudo update-ca-trust</h3>

2. Download and unzip the new installer (installer.zip) on all the nodes.
   Note: Give proper permissions to the folder by running sudo chmod 755 -R <installer-folder>.
3. Make the original cluster_config.json file available on the primary server.
4. Generate the new cluster_config.json file as follows:
   • If you have the old cluster_config.json file, use the following command to generate the configuration file from the cluster:

cd /path/to/new-installer ./configureUiPathAS.sh config get -i /path/to/old/cluster_config.json -o /path/to/store/generated/cluster_config.json

• If you do not have the old configuration file, run the following command:

cd /path/to/new-installer ./configureUiPathAS.sh config get -o /path/to/store/generated/cluster_config.json

5. Copy this cluster_config.json to the installer folder on all nodes.

Execution

Maintenance and backup

1. Make sure you have enabled the backup on the cluster.
2. Connect to one of the server nodes via SSH.
3. Verify that all desired volumes have backups in the cluster by running /path/to/new-installer/configureUiPathAS.sh verify-volumes-backup.
   Note: The backup might take some time, so wait for approximately 15-20 minutes, and then verify the volumes backup again.
4. Put the cluster in maintenance mode as follows:
5. Execute the following command:

   /path/to/new-installer/configureUiPathAS.sh enable-maintenance-mode

6. Verify that the cluster is in maintenance mode by running the following command:

   /path/to/new-installer/configureUiPathAS.sh is-maintenance-enabled

7. Create the SQL database backup.

Upgrade infrastructure on servers

1. Connect to each server via SSH.
2. Become root by running sudo su -.
3. Execute the following command on all servers:

   /path/to/new-installer/install-uipath.sh --upgrade --install-standalone-aicenter -k -i /path/to/cluster_config.json --accept-license-agreement -o /path/to/output.json

As part of this command, we take the backup of the Cluster state as well and pause all the further scheduled backups.

Note: This command also creates a backup of the cluster state and pauses all other scheduled backups.

Upgrade infrastructure on agents

1. Connect to each server via SSH.
2. Become root by running sudo su -.
3. Execute the following command:

   /path/to/new-installer/install-uipath.sh --upgrade --install-standalone-aicenter -k -i /path/to/cluster_config.json --accept-license-agreement -o /path/to/output.json

Execute rest of the upgrade on the primary server

1. Connect to primary server via SSH.
2. Become root by running sudo su -.
3. Remove redis section in patch json file.
   Because of a known issue in upgrade we manually need to edit the /opt/UiPathAutomationSuite/Installer/Upgrade/paths/patch.json file. This is fixed in 2022.4 version.
   To do so, open file using a text editor and remove the lines 22 to 25 and the ending , in line 21:

   vi /opt/UiPathAutomationSuite/Installer/Upgrade/paths/patch.json
   remove the lines: 
   {
       "upgradeScriptName": "redis_remove.sh",
       "upgradeScriptArguments": ""
   }
   remove comma on above line.

   After the changes the files should look as below:

   {
       "minSourceVersion": "<version_number>",
       "maxSourceVersion": "<version_number>",
       "minTargetVersion": "<version_number>",
       "maxTargetVersion": "<version_number>",
       "pre_infra_steps": [
           {
               "upgradeScriptName": "longhorn-single-to-multiple-replicas.sh",
               "upgradeScriptArguments": ""
           }
       ],
       "post_infra_steps": [],
       "pre_fabric_steps": [
           {
               "upgradeScriptName": "delete-immutable-jobs.sh",
               "upgradeScriptArguments": ""
           },
           {
               "upgradeScriptName": "rook-ceph-pre-1.7.9.sh",
               "upgradeScriptArguments": ""
           }
       ],
       "post_fabric_steps": [
           {
               "upgradeScriptName": "rook-ceph-post-1.7.9.sh",
               "upgradeScriptArguments": ""
           }
       ],
       "pre_services_steps": [],
       "post_services_steps": []
   }

4. Upgrade fabric by running the following command:

   /path/to/new-installer/install-uipath.sh --upgrade --install-standalone-aicenter -f -i /path/to/cluster_config.json --accept-license-agreement -o /path/to/output.json

5. Run the software installer by using the command below.

   First you need to connect to identity server and retrieve an installation token. Then copy this token and paste it in placeholder from the `cluster_config.json file.

   Once the token is there, you can run the AI Center install.

   /path/to/new-installer/install-uipath.sh --upgrade --install-standalone-aicenter -s -i /path/to/cluster_config.json --accept-license-agreement -o /path/to/output.json

   Note: This command disables the maintenance mode that you enabled before the upgrade because all services are required to be up during the upgrade.
6. After the successful upgrade and verification, resume the backup scheduling on the node by running the following command:

   /path/to/new-installer/configureUiPathAS.sh resume-scheduled-backups

Rollback on Error

Preparation

1. Create a separate folder to store the old bundles, and perform the following operations inside that folder.
2. Download and unzip the installer's old version (installer.zip) on all the nodes.
   Note: Give proper permissions to the folder by running sudo chmod 755 -R <installer-folder>.
3. Create the restore.json file and copy it to all the nodes.
4. Verify that the etcd backup data is present on the primary node at the following location: /mnt/backup/backup/<etcdBackupPath>/<node-name>/snapshots
   • etcdBackupPath - this is the same as the one specified in backup.json while enabling the backup's node-name;
5. node-name - the hostname of the primary server VM.

Cluster cleanup

1. Copy and run the dedicated script to uninstall everything from that node. Do this for all the nodes. For details, see the How to uninstall the cluster section in the Troubleshooting page.
2. Restore all UiPath databases to the older backup that was created before the upgrade.

Restore infra on server nodes

1. Connect to the primary server. This should be the same server node you selected during the upgrade.
2. Restore infra by running the following command:

   /path/to/older-installer/install-uipath.sh -i /path/to/restore.json -o /path/to/output.json -r --accept-license-agreement --install-type online

3. Connect to the rest of the server nodes one by one via SSH.
4. Restore infra on these nodes by running the following command on the server nodes one by one. Executing them in parallel is not supported.

   /path/to/older-installer/install-uipath.sh -i /path/to/restore.json -o /path/to/output.json -r -j server --accept-license-agreement --install-type online

Restore infra on agent nodes

1. Connect to each agent VM via SSH.
2. Restore infra on these nodes by running the following command:

   /path/to/older-installer/install-uipath.sh -i /path/to/restore.json -o /path/to/output.json -r -j agent --accept-license-agreement --install-type online

Restore volumes data

1. Connect to the primary server via SSH.
2. Go to the newer installer folder.
   Note: The previous infra restore commands were executed using the old installer, and the following commands are executed using the new installer bundle.
3. Disable the maintenance mode on the cluster by running the following command:

   /path/to/new-installer/configureUiPathAS.sh disable-maintenance-mode

4. Verify that maintenance mode is disabled by running the following command:

   /path/to/new-installer/configureUiPathAS.sh is-maintenance-enabled

5. Copy the restore.json file that was used in the infra restore stage to the new installer bundle folder.
6. Run volumes restore from the new installer bundle by executing the following command:

   /path/to/new-installer/install-uipath.sh -i /path/to/new-installer/restore.json -o /path/to/new-installer/output.json -r --volume-restore --accept-license-agreement --install-type online

7. Once the restore is completed, verify if everything is restored and working properly.
8. During the upgrade, scheduled backups were disabled on the primary node. To enable them again, run the following command:

   /path/to/new-installer/configureUiPathAS.sh resume-scheduled-backups

Offline Multi-node HA-ready Production Mode

Preparation

1. Identify any server (not agent) that meets the disk requirements for an offline installation. Referenced as a primary server throughout this document.

   If you are using a self-signed certificate, run the following command:

   <h3>Please replace /path/to/cert with path to location where you want to store certificates.
   sudo ./configureUiPathAS.sh tls-cert get --outpath /path/to/cert
   ### Now copy the ca.crt file generated in above location to trust store location
   sudo cp --remove-destination /part/to/cert/ca.crt /etc/pki/ca-trust/source/anchors/
   ### Update the trust store
   sudo update-ca-trust</h3>

2. Download the full offline bundle (sf.tar.gz) on the selected server.
3. Download the infra-only offline bundle (sf-infra.tar.gz) on all the other nodes.
4. Download and unzip the new installer (installer.zip) on all the nodes.
   Note: Give proper permissions to the folder by running sudo chmod 755 -R <installer-folder>.
5. Make the original cluster_config.json available on the primary server.
6. Generate the new cluster_config.json file as follows:
   • If you have the old ``cluster_config.json` file, use the following command to generate the configuration file from the cluster:

     cd /path/to/new-installer ./configureUiPathAS.sh config get -i /path/to/old/cluster_config.json -o /path/to/store/generated/cluster_config.json

   • If you do not have the old cluster config file, run the following command:

     cd /path/to/new-installer ./configureUiPathAS.sh config get -o /path/to/store/generated/cluster_config.json

7. Copy this cluster_config.json to the installer folder on all nodes.

Execution

Maintenance and backup

1. Make sure you have enabled the backup on the cluster.
2. Connect to one of the server nodes via SSH.
3. Verify that all desired volumes have backups in the cluster by running the following command:

   /path/to/new-installer/configureUiPathAS.sh verify-volumes-backup

   Note: The backup might take some time, so wait for approximately 15-20 minutes, and then verify the volumes backup again.
4. Put the cluster in maintenance mode as follows:
5. Execute the following command:

   /path/to/new-installer/configureUiPathAS.sh enable-maintenance-mode

6. Verify that the cluster is in maintenance mode by running the following command:

   /path/to/new-installer/configureUiPathAS.sh is-maintenance-enabled

7. Create the SQL database backup.

Upgrade infrastructure on servers

1. Connect to each server via SSH.
2. Become root by running sudo su -.
3. Execute the following command on all server nodes:

   /path/to/new-installer/install-uipath.sh --upgrade --install-standalone-aicenter -k -i /path/to/cluster_config.json --offline-bundle "/path/to/sf-infra.tar.gz" --offline-tmp-folder /uipath/tmp --install-offline-prereqs --accept-license-agreement -o /path/to/output.json

   Note: This command also creates a backup of the cluster state and pauses all other scheduled backups.

Upgrade infrastructure on agents

1. Connect to each server via SSH.
2. Become root by running sudo su -.
3. Execute the following command:

   /path/to/new-installer/install-uipath.sh --upgrade --install-standalone-aicenter -k -i /path/to/cluster_config.json --offline-bundle "/path/to/sf-infra.tar.gz" --offline-tmp-folder /uipath/tmp --install-offline-prereqs --accept-license-agreement -o /path/to/output.json

Execute the rest of the upgrade on the primary server

1. Connect to primary server via SSH.
2. Become root by running sudo su -.
3. Remove redis section in patch json file.
   Because of a known issue in upgrade we manually need to edit the /opt/UiPathAutomationSuite/Installer/Upgrade/paths/patch.json file. This is fixed in 2022.4 version.
   To do so, open file using a text editor and remove the lines 22 to 25 and the ending , in line 21:

   vi /opt/UiPathAutomationSuite/Installer/Upgrade/paths/patch.json
   remove the lines: 
   {
       "upgradeScriptName": "redis_remove.sh",
       "upgradeScriptArguments": ""
   }
   remove comma on above line.

   After the changes the files should look as below:

   {
       "minSourceVersion": "<version_number>",
       "maxSourceVersion": "<version_number>",
       "minTargetVersion": "<version_number>",
       "maxTargetVersion": "<version_number>",
       "pre_infra_steps": [
           {
               "upgradeScriptName": "longhorn-single-to-multiple-replicas.sh",
               "upgradeScriptArguments": ""
           }
       ],
       "post_infra_steps": [],
       "pre_fabric_steps": [
           {
               "upgradeScriptName": "delete-immutable-jobs.sh",
               "upgradeScriptArguments": ""
           },
           {
               "upgradeScriptName": "rook-ceph-pre-1.7.9.sh",
               "upgradeScriptArguments": ""
           }
       ],
       "post_fabric_steps": [
           {
               "upgradeScriptName": "rook-ceph-post-1.7.9.sh",
               "upgradeScriptArguments": ""
           }
       ],
       "pre_services_steps": [],
       "post_services_steps": []
   }

4. Upgrade fabric by running the following command:

   /path/to/new-installer/install-uipath.sh --upgrade --install-standalone-aicenter -f -i /path/to/cluster_config.json --offline-bundle "/path/to/sf.tar.gz" --offline-tmp-folder /uipath/tmp --install-offline-prereqs --accept-license-agreement -o /path/to/output.json

5. Run the software installer by using the command below.

   First you need to connect to identity server and retrieve an installation token. Then copy this token and paste it in placeholder from the `cluster_config.json file.

   Once the token is there, you can run the AI Center install.

   /path/to/new-installer/install-uipath.sh --upgrade --install-standalone-aicenter -s -i /path/to/cluster_config.json --offline-bundle "/path/to/sf.tar.gz" --offline-tmp-folder /uipath/tmp --install-offline-prereqs --accept-license-agreement -o /path/to/output.json

   Note: This command disables the maintenance mode that you enabled before the upgrade because all services are required to be up during the upgrade.
6. After the successful upgrade and verification, resume the backup scheduling on the node by running the following command:

   /path/to/new-installer/configureUiPathAS.sh resume-scheduled-backups

Rollback on Error

Preparation

1. Create a separate folder to store the old bundles, and perform the following operations inside that folder.
2. Download and unzip the installer's older version (installer.zip) on all the nodes.
   Note: Give proper permissions to the folder by running sudo chmod 755 -R <installer-folder>.
3. Create restore.json file and copy it to all the nodes.
4. Verify that the etcd backup data is present on the primary server at the following location: /mnt/backup/backup/<etcdBackupPath>/<node-name>/snapshots
   • etcdBackupPath - this is the same as the one specified in backup.json while enabling the backup's node-name;
   • node-name - the hostname of the primary server VM.

Cluster cleanup

1. Copy and run the dedicated script to uninstall everything from that node. Do this for all the nodes. For details, see the How to uninstall the cluster section in the Troubleshooting page.
2. Restore all UiPath databases to the older backup that was created before the upgrade.

Restore infra on server nodes

1. Connect to the primary server (which is the same as the one chosen during upgrade).
2. Restore infra by running the following command:

   /path/to/older-installer/install-uipath.sh -i /path/to/restore.json -o /path/to/output.json -r --accept-license-agreement --install-type online

3. Connect to the rest of the server nodes one by one via SSH.
4. Restore infra on these nodes by running the following command:

   /path/to/older-installer/install-uipath.sh -i /path/to/restore.json -o /path/to/output.json -r -j server --accept-license-agreement --install-type online

   Note: Run this command on server nodes one by one. Executing them in parallel is not supported.

Restore infra on agent nodes

1. Connect to each agent VM via SSH.
2. Restore infra on these nodes by running the following command:

   /path/to/older-installer/install-uipath.sh -i /path/to/restore.json -o /path/to/output.json -r -j agent --accept-license-agreement --install-type online

Restore volumes data

1. Connect to the primary server via SSH.
2. Go to the new installer folder.
   Note: The previous infra restore commands were executed using older installer, and the following commands are executed using newer installer bundle.
3. Disable the maintenance mode on the cluster by running the following command:

   /path/to/new-installer/configureUiPathAS.sh disable-maintenance-mode

4. Verify that maintenance mode is disabled by executing the following command:

   /path/to/new-installer/configureUiPathAS.sh is-maintenance-enabled

5. Copy the restore.jsonfile that was used in the infra restore stage to the new installer bundle folder.
6. Restore volumes from the newer installer bundle by executing the following command:

   /path/to/new-installer/install-uipath.sh -i /path/to/new-installer/restore.json -o /path/to/new-installer/output.json -r --volume-restore --accept-license-agreement --install-type online

7. Once the restore is completed, verify if everything is restored and working properly.
8. During the upgrade, scheduled backups were disabled on the primary node. To enable them again, run the following command:

   /path/to/new-installer/configureUiPathAS.sh resume-scheduled-backups

Downgrading Ceph From 16.2.6 to 15.2.9

For more information on downgrading Ceph, check the Downgrading Ceph from 16.2.6 to 15.2.9 from the Automation Suite guide.