Showing results for 
Show  only  | Search instead for 
Did you mean: 

Local cache on a citrix deployment


Local cache on a citrix deployment

Has anyone ever activated the local cache (in the MEGA work station options) on the Citrix application server?


In that configuration, the transactions would be stored on the application server until dispatch where they would update the repositories on the file server where the data is stored. One would assume this would reduce traffic between the application server and the file server, therefore increasing performance.


Perhaps this configuration would only provide negligible improvement, but I cannot think of any potential harm in setting things up this way. It requires having enough disk space on the application to store the data that will be kept there.


Do you think this is a good idea ? What could be a potential downside for this approach that I have not yet thought of?


Thank you for your feedback.

1 Reply

Re: Local cache on a citrix deployment

Hello Leslie


Local cache mode has been designed for WAN/LAN deployment.
I would not recommend activating local cache on the citrix server for several reasons


1) When local is activated transaction is not saved in the environment folder

This means than transactions files are not considered in physical backups of the environment folder


2) The local cache mechanism intends to limits the exchanges between a workstation and the file server when networks performances are poor.

When Citrix is involved, it is expected that the network performances between Citrix server and file server are very good (> 1 GBit bandwidth).


3) Activation of local cache is saved in a file (MEGAROOT.INI) that is profile dependant.

It would be heavy to configure each possible profile of the Citrix Server.


4) When working with with farm servers, it is not sure that a user will work on the same machine.

Let us imagine a farm of 2 Citrix servers and local cache mode configured.

  • User U1 logs it. His MEGA process runs on server S1
  • U1 opens a transaction
  • U1 exits MEGA (later): a local cache transaction is saved on a subfolder of S1
  • User U1 logs it. His MEGA process runs on server S2

--> U1 will not be able to resume his local cache transaction