Risk assessment?

[flambus]

is there a formula for determining the criticality of multiple server systems? is it risk assessment?

what is it.... links?

 

[Woodie]

Criticality? It sort of depends on what you're doing, and what you need.

I approach criticality from the business perspective: What is the impact ($$, lost business, damage to reputation,inability to complete mission, etc.) on the company if a particular application (functionality) is not available? What's the impact if the application is not available for: 1 hour, 2 hours, 8 hours, 1 day, 1 week?

If it's the application has different levels of availability (limited functionality), ask the same questions as above.

More details?

--Woodie

 

[flambus]

can criticality be broken into catagories?

a - high
b - medium
c - low
d - none


i work for a large company where different departments use different (and sometime the same) applications on different servers. payroll, supply inventory, blah blah....

my job is to validate all these systems and come up with a rating system for determining in what order do you bring a system back up from a failure.

 

[Woodie]

Heehee. We've just done the same thing.

We found that high, medium, low is too subjective. Have them rate functions by how long they can live without it.

Ie, payroll can be down for 1 day, A/R can't be down at all, A/P can be down for 1 week, etc...
Now create a matrix/db that shows where each component of an application lives (server names), and then set each server to have the same priority as the highest application component on that box (Max(application priorities)). Don't forget network infrastructure...if the router that the server is connected to is down, the server being up is no longer relevant.

--Woodie

 

[flambus]

that a great idea

this is the rating system they have in place. i think it stinks (you're right, too subjective)






Mission Critical ? An application that is vital to the successful continuance of the corporation. Without these applications, our very existence is in serious jeopardy. Recovery within 8 hours

High ? An application that needs to be recovered within 24 hours.
Medium ? An application that needs to be recovered within 48 hours.
Low - An application that needs to be recovered within 1 week.


Priority
A ? All Mission Critical applications.
B ? All High applications.
C ? All Medium applications.
D ? All Low applications.

thanks woodie

 

[Woodie]

We're a bit more serious. Our priority 1's tolerate 0 downtime. Full, automatic failover to devices in a different location. In some cases, includes fault-tolerant network as well (Yes, there are TWO networks, including different hardware, different circuits, and different cables, and different power sources).


flambus - YHEM

--Woodie

 

[flambus]

Thanks woodie