Cloning: method #1 (manual)
To clone a Cameyo server, follow these steps:
- On the image 'master' (the server you are about to clone), delete or rename Settings.xml and sk.dat in C:\RemoteAppPilot\Settings.
- Take the image.
- On the 'cloned' servers (the servers that were duplicated from the master), re-run the Cameyo installer and go through a regular registration.
Cloning: method #2 (automation)
To prepare a Cameyo server for cloning in a more automated way, proceed as follows. Preparation only needs to be performed once on the image master. Once your server is prepared, you don't need to do any additional manipulation on it going forward. It can still be used through Cameyo and can still be updated as a normal server. At any point you can then clone it.
- On the master server, edit C:\RemoteAppPilot\Settings\Settings.xml.
- In the Server node, add an attribute "cloneMaster" and set it to your master server's machine name. If in doubt, the machine name is indicated in Windows' environment variable COMPUTERNAME.
Restart the Cameyo service or reboot the server, to make sure the setting is properly entered. You should still be able to launch sessions on the master server at this point. If Cameyo sessions function properly on it, it means your server is ready for cloning at any point in time and as many times as needed.
- Once you've cloned a prepared master server, you should then simply change the cloned machine name. i.e. in the above example the clone machine should be renamed from "venus" to something else.
- Registration of the new server will then automatically occur for the new server.
- You will see the new clone in your Cameyo servers list, ending with a "(Clone)" in its name. Click on it, and adapt its settings accordingly, i.e. host name and ports.
Clustering and load-balancing
To group Cameyo servers together into a load-balanced Cameyo cluster, simply assign them to the same server cluster in your server's page:
When there are several servers assigned to the same cluster, Cameyo uses a weighted round robin algorithm to direct user sessions among available machines. This algorithm takes into account:
- CPU/RAM load (40%)
- Number of sessions (40%)
- Distance from user (20%) (negligible if servers are within the same area)