Vault content availablility.

[SM_12968661]

Dear WC adminstrators,

I do have some ambiguity on below scenarios how the content is managed.

1. Is content copied or moved from cahce vaults if Cache vaults and replica vaults are set up in replica site, when any user uploads a file connecting to any replica site?
2. Is content moved or copied when file initially uploads to cache vault and then goes to master vaults incase of master site?
3. How is content accessible by user B connected to main,if user A uploads the file to replica site.
4. how the content is accessed by an user in Replica site B if an user from Replica site A uploads to Replica A site vaults,
5. if a user uploads a file to main, how is this available to userB connected to any replica?
6.  

[BenLoosli]

Haven't done this in a few years but from what I remember back then.

Files are replicated between the replica servers to the main server at a predefined interval set when the replica server is configured. You set this depending on your needs.

Files are copied between the replica server and the main and the main to the replica. There is always a 'master' file in the main server, except new files in the replica saved between sync times.

The Windchill database knows about all files in the main and all replica servers. If a file is not available in the replica server, Windchill will 'transfer' a copy to the local replica server.

You can pull files directly from the main server while working at a replica site. When saved, the file will be saved on the replica server.

As far as cache vaults go, I believe that is always a copy. One of the reasons you need to delete unreferenced files periodically is to clear disk space used by files that have been copied to the main vaults.

[SM_10838685]

thanks for the comments.

if content from main site to replica gets copied, the size of the master vaults and replica site vaults should be same, which is NOT in most of the cases.

I understand the replication scheduler does the job of replicating the content from replica site to master.

My question is how a file available to the other replica site user, when the file uploaded from another replica site? also, if a file upload to main sites, how is it available across all the replica sites?

[BenLoosli]

A lot depends on how you have the replica sites set up. When I did it, we only put certain contexts into the replication, not everything as that site did not need all of the files. 

Windchill knows what files are on which replica server since the meta data is ALWAYS loaded into the master system. A request for a file on Replica Server A from a user on Replica Server C would initiate the transfer from Server A to the Master and then to Server C. There may be some delay but should not be detrimental to the work flow.

Files on the master are synced with the replica servers at a defined interval depending on contexts available to each replica server. If all sites work on all files, then there should be close relationship to file sizes on all servers. That may not be true when dealing with cache files and published viewables and thumbnails, which might only be on the master server, explaining some of the size difference.

[jlecoz]

About file vault replication there are mainly three related processes:

• Synchronizations - Moves files from cache to master vault.
• Revaulting - Moves file from master vault to master vaults
• Replica - Moves files from master vault to replica vault.

• Sync is scheduled, but it can be triggered when a specific task occurs. For example, working on a replica site, you check in the CAD doc. It goes first in the cache, but then the visualization service starts to generate the viewable, requiring a sync from the cache to the master.  Then the viewable generation can be done on the main server environment. Another reason is if another user working on another site wants to download the file, it is then synced and the master to made accessible.
• Replica - you may have a different replica setup and rules. Depending on the setup, you may not have all the files replicated. For example, you can set up replication on demand. In that case, only the file that is requested by the user to download will be replicated for a specified amount of time. And you don't have a 1-1 count between the master and the replica. This is a setup I sometimes used to simplify replication management, removing the need for replication rules. So first check the replica rule setup to understand the vault size difference.