Instructions for updating a multi-node environment for production running NoMachine (v.5)
|
Considerations - Pre-requisite is to download the server packages labelled as 'UPDATE VERSION' from your Customer Area. - Updating the installed software requires the termination of all the running sessions. These sessions cannot be recovered later. - The cleanest way to upgrade a multi-node environment is to (i) initially update installation on the nodes, then (ii) update installation of the main server or of the cluster servers when the failover cluster is enabled. |
||
| Please read carefully all the steps below before proceeding with the upgrade! |
1. Planning server downtime period
Updating the installed software requires the termination of all the running sessions. These sessions cannot be recovered later. This is necessary for installing the new libraries and binaries. Processes already loaded in the system memory lock down the corresponding binaries and libraries that cannot be otherwise replaced.
The update procedure implies a shutdown of the server.
As a precaution, we suggest to:
- Send a broadcast message to all connected users informing about ongoing operations:
$ sudo /usr/NX/bin/nxserver --broadcast "Your message here"
- Shutdown the NoMachine server
$ sudo /usr/NX/bin/nxserver --shutdown.
NOTES
- Stopping the remote nodes before proceeding with the update is not required.
- If there are two servers in failover cluster mode: shutdown NoMachine server on the primary server and on the secondary server.
2. Update the remote nodes
Proceed to update the NoMachine installation on each of the remote nodes.
3. Update the server(s)
Proceed to update the NoMachine server installation.
The update procedure takes care of starting all the NoMachine services, it's not necessary to start the server(s) manually, unless the failover cluster is enabled.
NOTE
- If there are two servers in failover cluster mode:
(i) update the primary server and the secondary server. Order of update is not relevant.
(ii) Then restart both cluster servers after the update to make sure that the cluster interface is created on the primary server:
$ sudo /usr/NX/bin/nxserver --restart
TROUBLESHOOTING: What to do if the server installation has been upgraded before the nodes
- In case of a multi-node environment without failover cluster functionalities, run the following command on the main server host to retrieve resources from the nodes:
$ sudo /usr/NX/bin/nxserver --nodedbupdate
- If the cluster failover is enabled instead:
(i) run the following command on the primary server:
$ sudo /usr/NX/bin/nxserver --nodedbupdate
(ii) and then synchronize the cluster dbs by running on the primary server:
$ sudo /usr/NX/bin/nxserver --clusterupdate