Enterprise Messaging Notes

We, like many, are eagerly testing and evaluating Domino 8.5, DAOS and prerequisite transaction logging (TL).  We have not run TL before, mainly based on things like very long crash recovery times in the r5 days with TL..  We have an environment running Veritas Net Backup (NBU) 6.5.4, and significant business needs for point in time restores.  So far, those test point in time restores are behaving rather oddly, and taking forever to parse through the transaction logs and roll forward/roll back, one recent test for one restore took more than 20 hours, after "tuning".  Veritas, as per usual, is less than helpful, and the DAOS backup strategy (NSF backup and file system backup for the NLO's) will not be integrated into a coordinated offering from Veritas until 7.x of the NBU client for Lotus.   IBM support has punted, appropriately, in helping us understand the odd behavior with NBU.

It has me wondering.  Disk is cheap, cheaper all the time.  Although the advantages of saving disk space seem like good goals to shoot for, is the additional complexity of having to do transaction logging, and all the seeming complexity that comes with that for restores-- essentiually relying more heavily on your backup vendor to fully integrate and coordinate the aspects of recovering Domino data-- worth the effort?  Veritas cannot give any method in the current version to coordinate the 2 backups, so our Unix admin is scratching his head trying to figure out how to know when to schedule the file-system NLO backups, that need to occur after the NSF's are backed up.  All of this creates an increased dependency on backup vendors and products, simply to do the routine tasks like file restores that were pretty-much not dependent on them prior to DOAS and TL's, and in our case with Veritas, that is a dependency that is not attractive.

We're also starting to look at replacing NBU, possibly with something like Tivoli for backups, which as an IBM offering, is more integrated with DAOS. Now we're considering changing all of that, simply to de-dup attachments, which is another huge set of things to do.

We strive for simplicity and to avoid additional cost, complexity and increasing dependencies on "utility" software products and vendors to manage and leverage our Domino environments, and although decreasing storage seems great, I'm starting to think DOAS may not be worth adding more complexity, and headaches for doing simple things like point in time restores, and/or changing all the things that come with backups.

Interested in any comments you may have on this.  Thanks in advance.