3.4 folder replication, is it practicle?

[jruffsr]

I have assumed responsibility for Intralink 3.4 with a central server in US and a remote vaults in England and China with a couple of more on the horizon. We have common hardware and components on the central server which causes huge performance hits when a user access an instance [many of which are members of very large family tables].

I have read about replication of these objects but I would like to hear if anyone has done this and how was the performance increase [if any]. What methodolgy did you use to sync these?
Thanks and have a great day.
Jeff Ruff

 

[FishNut]

There are two type of replication, Package and Cluster. I have a remote site in another state. We use cluster replication. We have common hardware parts like you described. These folders are replicated on a nightly basis. There is a script or job that runs every night at the remote site that replicates any changed objects in the folders designated. I did not set up the replication, or its job script so I cannot be specific about it how to do it. It has worked for us for about 7 years now.

 

[proengineertips]

Using replication will save the checkout time (local copy) for pro/E
user. The system is retriving the files thru LAN after applied
replication; no longer limit by the internet speed.

 

[tundra]

RE: Cluster replication


It's my understanding with IntraLink 3.x and cluster replication, upon CHECKINGIN files retrieved from a replicated cluster, these same files and their metadata go back to the original file server & data server. So for cluster replication across a WAN, "check-out" is "fast," but "check-in" is "slow."