On TechRepublic: Weirdest error messages of all time
BNET Business Network:
BNET
TechRepublic
ZDNet
TalkBack 1 of 19:
Next »
Off site backups, multiple locations required
No matter how robust the on-site backup, there will always be site specific failures. No matter how superficial the document is, you must independently check that the procedures are updated and followed. How many nicely written descriptions get totally undermined by not being followed, people get lazy, don't fully understand and don't care.

They need to do the following:

1. Get several best of breed procedures and hire two separate outside people to review the systems and write up separate updates along with a version by the internal IT staff. Have an open process to make sure that everything is covered from all of the versions ... if there is a area of difference, pick the best or if in doubt bring a well-known expert.
2. You must have external backups. You must keep full versions for periodic date going back each quarter and then each year for at least 5 years. Why: Any data center can be comprised by sloppy procedures, undetected corruption of data in a production systems the may not show up for months, accidental user data deletion, incompetence, power systems failures - even UPS systems can "meltdown", flooding, earthquakes, physical assault, fires, and maybe something you did not foresee. Remember, some backups might send someone to jail ... that's a lot of incentive.
3. The offsite backups should be alternated amongst two or more separate locations. This adds little cost. Access to these backups (and recycling of existing backups) must get an independent signoffs from those who do the backups. Separate signoffs for each site at a high level ... needing the backup should have very high visibility in the organization ... the onsites should have been sufficient and you can't afford to lose an offsite backup in this case.
4. Verify the backup. An independent party should restore the system and make sure it is worth any! Not just the data but the systems as a whole.

If the data cost billions, you can spend millions. It is hard to imagine a company surviving a massive data loss.
Posted by: radolan@...   Posted on: 08/18/06 You are currently: a Guest | Members login | Terms of Use

Alert moderator to an offensive message

Subscribe to this discussion via Email or RSS

Off site backups, multiple locations required  radolan@... | 08/18/06
Worst case.  Anton Philidor | 08/18/06
The great blackout of 2003  Roger Ramjet | 08/18/06
This is not actually a problem...  bportlock | 08/18/06
Normal risks and workarounds  dave.leigh@... | 08/19/06
Let's say...  Anton Philidor | 08/19/06
Well, obviously.  dave.leigh@... | 08/20/06
I shut down - a client dieselled through  murph_z ZDNet Moderator | 08/20/06
One statement...  slopoke | 08/18/06
aah, but there's a miracle .. wave hands and say  murph_z ZDNet Moderator | 08/18/06
Where do you live?  dave.leigh@... | 08/19/06
Agreed - however, if you're big enough  murph_z ZDNet Moderator | 08/19/06
And if the "redundant mirror"...  Anton Philidor | 08/19/06
No mechanism  murph_z ZDNet Moderator | 08/20/06
Your quote did mention testing...  Anton Philidor | 08/21/06
Oh my! Whatever will we do with our data if the planet's destroyed?  dave.leigh@... | 08/20/06
Well, sure.  dave.leigh@... | 08/20/06
Backup  TonyMcS | 08/20/06
Some invalid inference  jorwell | 08/22/06

What do you think?

SponsoredWhite Papers, Webcasts, and Downloads

advertisement
advertisement

SmartPlanet

Click Here