Environment Sizes

[jgorman]

General question about EV environment sizes. My environment is 60,000 users with just over 101TB of archived mail. How does this stack up with others? 

[GertjanA]

Hi jgorman,

That is really little info to compare :)

How long have you been running archiving? Any retention? How is your sharing configured? Which version of EV did you start with, and what version do you have now? All active mailbox-archives, or do you also have many ' leavers'.

Anyway. I've seen a few large environments in my region.

1 - 140.000 archives, of which about 45.000 inactive. = 95TB (many many active archives that are really small (low mailusage). Started using EV in 2004

2 - 25.000 archives, of which about 16.000 inactive = 56TB (started using EV in 2003)

This obviously excludes Journal Archives. Both examples above had 'keep items forever'.

[jgorman]

We started with EV 6.0 and are now on 10.0.4. Single VSG, retention is Forever. All active mailboxes are archived and we see turnover of about 1200 mailboxes/archives a month (leavers). Once users leave we remove their archives after 60 days. My curiousity is on two fronts...1. How many sites are out there over 100TB and 2. How or ARE those being backed up?

[GertjanA]

wauw. so even with removing old archives you still have that many.

The largest I've seen was using collect and migrate. The open partitions are 500GB, collection runs to create cabfiles 30 days after archived, size = 10MB. The migrate puts the cabfile after 5 days on an HP EVA SAN server, which replicates data natively to a DR EVA. Backup is being made of the open partitions only, and indexes ofcourse. If both EVA would crash, 'houston we have a problem'.

Another I've seen has backups of closed partitions. The open partition is backed up using TSM. When closed, it is backed up for another 14 days (to allow collection to run, no migration). After 14 days, the open partition is backed up only. As this environment has no expiry, nor deletion, old backups are kept (as the data does not change) to be able to restore environment if necessary.

I also have heard of environments that use Centera's, NetApp's or HCP devixes for storage. HCP I know for sure, does blockbased replication with hashing. That takes away the need for backup. I also know that if you use Quadrotech EVNearSync, you don't need backup (also hashes). Both indicate having a DR environment.

With such a large environment you have, and having large turnover and deleting of archives, what storage do you have?