Click to See Complete Forum and Search --> : Disaster Recovery Planning, Contingency Planning


Harvey
03-29-2001, 10:04 AM
Are the same thing... I wanna find some information about this stuff.... anyone know of any good sites or books?

I have checked these out... http://www.contingencyplanning.com/ http://www.drj.com/

but they still leave me feeling unsatisfied... need more stuff.

anyone know of any good sites or references to this topic?

bdg1983
03-29-2001, 07:37 PM
For Linux?

Seem to remember seeing the topic at www.linuxdoc.org (http://www.linuxdoc.org)

Harvey
03-29-2001, 08:16 PM
Actually just in general. Though I suspect the topic usually applies to *nix enviroments cause usually the planning is for huge servers and things.

bdg1983
03-30-2001, 08:41 AM
Since we use Comdisco for our disaster recovery, perhaps http://www.comdisco.com/ has some info you can use.

There maybe books at Informit (http://infobase.informit.com/) that could help. Free registration.

posterboy
03-30-2001, 01:24 PM
Here's how I do it, if worth anything to you. I have no real faith in anything I am doing to prevent a disaster. I don't believe in "unbreakable" firewalls, and I don't believe in my 'puters integrity. I have seen HD's trashed by other hardware issues cropping up. For 6,8 bucks, you can buy removable hard drive trays. I change drives each month. I always have a drive disconnected in the desk drawer, from 1 to 30 days behind my current one. There is a semi-permanent drive also, and scripts move "stuff" from the main drive to that one at changeover time. Change, then scripts bring updates into the new drive. If disaster, I can clone the desk drawer drive in a few minutes, and I am going again. I can lose "some" data this way, but it's not a disaster, really. I also do backups oif important stuff weekly, so that if i lose it, I can pull that stuff into the freshly cloned drive. I AM convinced that the black-hats CANNOT hack the drive that's in the desk drawer. The house burning down is about all I haven't got covered in this way. Ray

tnordloh
04-06-2001, 07:30 AM
This is mostly for those of us who aren't exposed to server environment. Here's what you can typically expect.

Here's the setup at the server farm where I work work;
Out biggest client (A cellphone company - you would recognize the name) owns one of the servers. We pay something like 30,000 a month for HP to maintain an N-4000 in Florida, so that if there were a natural disaster the sysadmin could jump in a plane with a pair of hard drives (or backup tapes, if the server is that bad off) and pop 'em in that N-4000. I think it's located in Florida, don't quote me on that though.

Our computer is housed in a special room, for which HP has specific requirements, such as protection from grounding, specific ups steups, accessibility for the wheeled units they use to carry these servers around on, and so on. If these protections aren't in place, hp charges exorbitant fees to maintain the boxes. more than this speciallyu built room would cost, from what I gather...

The N-4000 itself houses the actual cpu, and a pair of mirrored hard drives. Recently our sysadmin upgraded to a secondary disk array containing 4 scsi hard drives. 2 of these drives are exact mirrors of the other 2, and their main purpose is to provide 100% uptime, and to allow the sysadmin to roll back os upgrades in minutes. He can literally walk into the operations room and swap the drives around, therefore allowing him to reboot from a previous configuration in seconds, not counting reboot time. All this is for storing the os. We haven't even gotten to data storage yet...

The actual database information is stored in a gigantic box. It's called an XP256 and it's literally that. A big cabinet full of hard drives.

The building itself has an outside powerline, as well as a backup generator that runs off natural gas. We could literally drop off the power grid indefinitely, until that gas stops a-flowing. There is also a contract arranged with a local generator company who will provide us with a diesel generator within half an hour or so.

The backup scheme is twice a day. Once at noon, once at midnight. They have what's called a business copy set up in the xp256 which they unsync for about 2 hours, allowing them to do a full backup. This costs about 2 minutes of time while the system pauses during the drives' unsyncing process. Believe me, if that system isn't back onlie in less than 5 minutes, phone calls start rolling in. The backup itself runs off a 2/15 tape jukebox (2 drives and cassettes holding 15 tapes). The tapes are stored in a secure rented facility that is underground about 10 minutes away from ATCIS. It's locked up; I have a little card in my wallet that tells them I'm allowed to get tapes there. As far as security problems, that is the most likely source of info loss. While I'm not impressed by their security, I'll leave out a description of what I see as their faults. The tapes themselves are reused on a 60 day rotation schedule, with the exception of monthly and yearly backups which are archived for several years. To give you an idea of the cost of this, these tapes cost 80-100 dollars apiece, and each backup requires 2.

All this just so some pimple-faced kid in a mall kiosk won't miss out on the chance to sell a dam cellphone cause the computer is down.

From this, I think I've learned that the most important thing for disaster recovery is money. Right from the start, buying that extra hard drive costs what a decent ram upgrade would. That's the simplest backup, and right there you're looking at spending a hundred bucks or so.


By the way, the N-4000 isn't even the top-end machine at HP. They have something called a SuperDome, I have no idea what it's like. I've just heard my sysadmin talk about it like the holy dam grail. Can you imagine the money being laid out by those who are maintaining those things?

So what are some of the ideas for maintaining data?

1. Expensive computer systems waiting to be used
2. Separate hard drives for the database and the os
3. simple but expensive schemes for rolling back os upgrades and patches
4. simple but expensive tape rotation procedures.

Harvey
04-06-2001, 04:30 PM
That's exactly the type of information that I was looking for! Very very interesting.

Tell me more! :D

I'm going to be doing an internship with large company this summer and they are going to have me doing Diaster Recovery Planning, yet I hardly know anything about it yet.

This information really helps me out! Thanks a lot man, and keep it coming if you have any more stories.