• jackimo - Friday, January 19, 2018 12:09 PM

    Jeff Moden - Wednesday, January 17, 2018 10:41 AM

    I guess my recommendation would be... to educate the people involved as to what RPO and RTO actually means and what the impact to the business would be for (for example) the rather miserable 24 hour RPO that they currently seem to have and the fact that there might be no RTO possible unless someone actually verifies that the backups are restorable.

    As for Dev and Staging boxes, a lot of people are working their butts off to get code developed and tested.  For that reason, I treat both boxes just like the Production box and have an even tighter RPO and RTO because things happen and people DO accidently delete stuff (data and or objects) in the "heat of battle".  To have lost a two day test setup would be stupid when something is so easy to recover from simply by doing the right thing with backups and testing restores.  And, yes... I restore my two largest databases every night right after the backups complete.  Granted those two databases only take up about 2TB of space but it's essential that we be able to quickly restore if all the other fancy methods that the infrastructure team has (very well) put together.  It also gives me a regular measurement of precisely what the RTO will be in case of such an event.  Even my 2TB telephone system (we're required to keep call recordings forever) has been setup for an otherwise insane "get back in business time" of only 6 minutes and a total RTO of only a few hours (for the remaining 8 years of calls) because of the way we have backups constructed, which is also a strong justification that "one size" DOES NOT fit all.

    Thanks for the reply Jeff.  In our organization I don't generally communicate directly with the business units.  I think if I contacted them directly it wouldn't go over very well.  I think I'm close to having control of backups but the share restrictions just make restores difficult.  And in my mind the restore is the most important part.  I think I mentioned that in the 7 months I've worked here there has not been 1 restore using our enterprise backup appliance.

    Yowch.  I don't envy that position at all.  Hang in there and keep trying for what's right.  Hopefully they won't need to find out if a restore actually works between now and when they finally come to their senses.  My hat's off to you for your perseverance.

    --Jeff Moden


    RBAR is pronounced "ree-bar" and is a "Modenism" for Row-By-Agonizing-Row.
    First step towards the paradigm shift of writing Set Based code:
    ________Stop thinking about what you want to do to a ROW... think, instead, of what you want to do to a COLUMN.

    Change is inevitable... Change for the better is not.


    Helpful Links:
    How to post code problems
    How to Post Performance Problems
    Create a Tally Function (fnTally)