Draw a time line on an X axis.

In the middle, put a big red D.  This is your disaster.
Somewhere to the left, put a P.  That is your RPO.
Somwhere to the right, put a T.  That is your RTO.

As the time between P and D decreases, or as the time between D and T
decreases, or both, your price and complexity increase.  Quickly.

A propoer business impact will put a $ value to the loss of data or
productivity, per system.  The answers to the questions come from the
business, not IT.  How much $ is at risk if the carpet spec sheets are not
available for 1 hour, for 2 hours ....  You then factor in probability of
occurance.

If the spec sheet is unavailable for 4 hours means you lose one sale.  That
one sale is $1000 in revenue.  Probability of that "event" occuring is 1x
every 2 years.  That means (simplistically) you should spend no more than
$500/year to protect the carpet spec sheets.  Again, very simplistic, and
you would repeat for each system to help drive the overall requirements.

I would spend a few thousand dollars with a business impact consultant
first.

On Tue, Apr 6, 2010 at 10:52 AM, John Aldrich
<[email protected]>wrote:

>  Ah. Looked up the two terms. No, I have not really done an “in-depth”
> formal investigation of those, but I would say that I really don’t want to
> lose ANY of the data, period. A recovery time of a few hours should be
> acceptable, but I’d certainly like to keep it under 1 business day. After
> that, people are going to start getting impatient, I believe, for access to
> certain business-related files (spec sheets for carpet, etc.) As previously
> mentioned, I also want to have our email store on the storage appliance, and
> while I’m sure our ISP would be more than happy to act as the backup email
> host, it would be better if we didn’t have to keep our email on their
> servers any longer than necessary. In addition, if I back up user’s desktops
> to the network (most likely using folder redirection) they might get a bit
> “antsy” if they don’t have access to their files. J
>
>
>
> So, no, no FORMAL RTO/RPO, but I don’t really want to lose **any** data
> and I want to restore functionality as soon as humanly possible. I don’t
> want to take a week or more to get a new SAN installed if I can possibly
> help it. That’s too long. I would say up to 24 hours (including weekends, if
> necessary, but preferably not!) would be the maximum “down time.” Most of
> our operations run on the AS/400, so there would be little business-critical
> data on the Windows side of things, but a lot of business-IMPORTANT
> information that would be hard to replace.
>
>
>
> [image: John-Aldrich][image: Tile-Tools]
>
>
>
> *From:* Kevin Lundy [mailto:[email protected]]
> *Sent:* Tuesday, April 06, 2010 10:37 AM
>
> *To:* NT System Admin Issues
> *Subject:* Re: SAN question
>
>
>
> John - have you completed a business impact analysis?  Derived RTO and
> RPO?  I think that will go a long way towards helping define technical
> options as well as the finances.
>
>
>
> To try to design the technology first, in my opinion, is putting the
> proverbial horse before the cart.
>
>
>
> Kevin
>
> On Tue, Apr 6, 2010 at 10:28 AM, John Aldrich <
> [email protected]> wrote:
>
> No, I have not narrowed it down. I initially hoped to get the project done
> this year, but the economic downturn has hurt our cash flow such that we are
> in a “holding pattern” on any “discretionary” spending and the SAN project
> certainly falls into that category.
>
>
>
> You bring up a good point about what happens if you only have one SAN and
> it goes down. That’s one reason I was thinking about having a “D/R” SAN at a
> remote location. We do have a remote location about 30 minutes away. Another
> option would be to co-lo the D/R SAN with someone. I just got off the phone
> with a Global Crossing rep who wanted to talk to me about things they could
> do for me, especially in the way of conferencing, etc. We got to talking
> about the IT priorities for the next couple years and I mentioned the SAN
> project and asked if they provided online backup services. She said that the
> only thing they could offer would be co-lo space and that got me to thinking
> that maybe that might be the best of both worlds… co-lo the D/R SAN at a
> professional co-lo facility.
>
>
>
> [image: John-Aldrich][image: Tile-Tools]
>
>
>
> *From:* Martin Blackstone [mailto:[email protected]]
> *Sent:* Tuesday, April 06, 2010 10:14 AM
> *To:* NT System Admin Issues
> *Subject:* RE: SAN question
>
>
>
> If you are just backing up to tape and your SAN goes down, where will you
> restore too? Do you have a spare disk pool to use?
>
> If it was “me”, I would be looking at a SAN solution that offers its own
> proven DR solution.
>
> Since I only know NetApp, they have a tool called SnapMirror that is built
> into the OS. You pay for the license and plug in the serial.
>
> Then setup your DR targets and let it rip. If your primary SAN goes down,
> you can do some clicks and bring the system online with all your data ready
> to access.
>
>
>
> But you seem to be talking about a lot of things you want. You want DR, you
> want clustering. If you cluster, maybe you only need to backup to tape.
> Unless you want to buy a clustered SAN and a DR SAN. Of course if you are
> going to have a DR SAN, I assume you have a DR location?
>
> I mean if the building burns to the ground do you have a location with the
> resources needed to keep the company running? Not just hold the data?
>
>
>
> Have you narrowed this down to 3 vendors yet?
>
>
>
> *From:* John Aldrich [mailto:[email protected]]
> *Sent:* Tuesday, April 06, 2010 6:52 AM
> *To:* NT System Admin Issues
> *Subject:* SAN question
>
>
>
> Guys, I’m still working on my storage needs, as the project I’ve been
> working on probably won’t get approved until early next year at the
> earliest. I was talking to a D/R consultant recommended by one of the folks
> on this list. Unfortunately, he does not work with SMB clients, only large
> clients such as Coca Cola, etc.
>
>
>
> I had been thinking of getting two SANs and having one replicate to the
> other for D/R purposes. Most of our operations run off the AS/400 so that
> would not be much affected (except if we are able to some how back up to the
> SAN, which is unlikely with our current AS/400, due to disk space
> limitations on the 400) one way or the other by the SAN project. The
> aforementioned consultant suggested that we look into getting just one SAN
> and a tape backup for it or online backup service instead of doing two SANs.
> Most of the data on the Windows side of things would be hard to replace if
> it died, so while it’s not “critical” to our operations, it’s still highly
> important.
>
>
>
> What do you guys think of that suggestion? Would any of you guys do
> something like that? Why or why not?
>
> Also, anyone know any D/R consultants in the North Georgia area who work
> with SMB clients?
>
>
>
> [image: John-Aldrich][image: Tile-Tools]
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>

~ Finally, powerful endpoint security that ISN'T a resource hog! ~
~ <http://www.sunbeltsoftware.com/Business/VIPRE-Enterprise/>  ~

<<image001.jpg>>

<<image002.jpg>>

Reply via email to