Data replication between sites


I am trying to figure out the best way to replicate User data between 2 sites.  Such that:

Need to have:
When a User logs in to site A (via RDP or Citirx) GP maps them to a Local share with their data.  They make a change and that change is instantly (block Level) replicated to site B.

Nice to have:
When a user Logs in to site A (rdp/ctx) and opens and works on a document and a Meteor hits the Data Center.  The user is able to log in to Site B and recover from auto save....

Is this possible?

I was thinking I could use DFS-R but the scalability doesn't seem to be there as we are talking about 50,000 users.  

And SAN Side replication doesn't seem to talk much about this on their Websites (EMC)

Note: I left the question Vague as to not color your responses with my preconceptions.
Who is Participating?
Duncan MeyersCommented:
From what you'e told us, synchronous SAN-based replication is what you're looking for. EMC has:
For Symmetrix/VMAX:
RecoverPoint/SE synchronous or RecoverPoint synchronous

For VNX, CLARiiON, Celerra:
RecoverPoint/EX synchronous

If you can afford to be slightly behind, asynchronous replication will save you a bucket o' cash.
For Symmetrix/VMAX:
RecoverPoint/EX  or RecoverPoint

For VNX, CLARiiON, Celerra:
VNX Replicator

If you have sufficient WAN bandwidth, asynchronous and near-synchronous RecoverPoint can keep the DR copy within one or two transactions of the source. Synchronous RecoverPoint keeps the DR copy in complete sync, but you need lots of bandwidth and your data centres need to be within about 100km of each other.

Another technology to take a look at is VPLEX Metro. VPLEX Metro allows you to stretch a Microsoft (or other) cluster over distance with zero downtime failover. Seriously cool stuff.

There's also integration between MirrorView and Microsoft clusters with MirrorView/CE - this also allows you to do a stretch cluster.
Please quantify the amount of data that changes during peak time, for all users at both sites.

Also, what kind of lag time are you willing to live with vs. how long a delay till the site gets replicated.   Is this supposed to be realtime, or if data wasn't replicated for 5 mins or 5 hours, does it matter?

For a moment, we need to avoid the software, we have to look at the physics and  understand worse case scenario and bandwidth constraints.
Duncan MeyersCommented:
Thanks! Glad I could help.
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.