Save
Replication, its good stuff and we have talked about it on this blog many times but there is still a little confusion out there about the meaning of the terminology and how it behaves.
Replication comes in two flavors, File Store replication and Database replication (used in Connected Workgroups).
File Store replication relates to the transfer of the actual files from site to site either on demand or on a schedule.
This file transfer schedule is configured by the Vault administrator in the ADMS Console and most companies will set this to run once a day prior to backup on the main site outside of working hours. You could of course set this up to run every hour or more in rapidly changing and highly collaborative design environments but typically having the latest files on site at the start of each day is sufficient and users can always transfer a file "on demand" if they need it immediately (You don't have the file locally – do you want to copy it now?).
Database replication is where the SQL servers transfer meta-data back and forth between SQL servers – this is a process managed by SQL (not by Autodesk Data Management Server), you do not set up schedules and although you can tweak SQL replication settings its typically not a great idea to do this unless guided by Autodesk.
Given a healthy connection SQL replication should occur every 2 minutes, the important point here is that users should be seeing file status information as close to real time as possible to accommodate remote collaborative design.
In addition to this, File Store replication can be configured to only replicate specific Vault files to reduce network traffic, although you need all files at the publisher site for backup you can ignore files not required for use at certain subscriber Workgroups. Database replication however replicates all meta-data all the time. Vault security should be used to prevent users from accessing restricted files or data in this case.
These 2 types of replication need separate monitoring and are logged via separate methods (See general log information here and SQL monitoring here).
In terms of their use – Filestore replication can be configured to work alone, this is the case when two or more sites are well connected and share a single SQL database, improving user productivity by making their files available locally. Connected Workgroups (or fully replicated environments) use both file store and database replication – typically due to poor connection between the sites meaning a single SQL database is hard to support.
Read more on the WikiHelp links above to better understand and hopefully simplify the support and maintenance of your replicated environment.
