Joined: Sep 16, 2006 Posts: 3590 Location: London, UK
Posted: Wed Feb 06, 2008 12:11 am Post subject:
Why would you give that information to a customer
1 - they will hold it against you
2 - they will hold it against you
3 - they will hold it against you
I dont think you did the number correctly
I did the server * etc acrocss and came up with 89.4319 %
Availability is just not raw numbers. It needs to be done in context to the SLA
The working hours for the tool / service you are using is 0800 - 1600 local business time Monday to Friday. The service is expected to be available 95 % of time or .95 of 8 hours per day for a 5 day period.
So that would mean 5 % of 480 minutes which is 24 minutes per day of potential service unavailability or 2 hours per week.
The SLA would determine whether the working hours local includes holidays etc for the local as well as global holidays if the server is not in the region where the user is
That is for the customer
For you as AM and taking your example you need to adjsut you calculation areas a bit
Start with the application supplying the service
Is it on 1 machines (Single point of failure)
or is it spread over 10 machines in a cluster. At what point is the #-
of servers being unavailable have an impact on the application and users
Is the network connections from the server / server farm a SPOF or are there redundant links from the server farm to the network so that users can access the equipment.
Look at how the network is structured - router pairs, switch pairs
So then look at the the users of the service
Can the service be available but users can not get to it ? Yes
The Web site / servers are up 100% availability but MY PC can connect because I have a bad Network card. I have an availability issue not the service
In order to have AM and to have it usefull for you and your customers, Service level management and the definition of the service is requried _________________ John Hardesty
ITSM Manager's Certificate (Red Badge)
Change Management is POWER & CONTROL. /....evil laughter
One way of looking at it may be that there is a link showing as "link (97%) "
So this is the weakest "link" in the chain provided. This may be the "minimum" availability of this "service" when taken the other %'s into consideration and assuming that this part of the chain has no failover etc. Maybe the question asked meant this instead of optimal.
Just presenting a "possible" answer to the question. _________________ Mark O'Loughlin
ITSM / ITIL Consultant
Joined: Jun 18, 2007 Posts: 10 Location: Adelaide, South Australia
Posted: Thu Apr 15, 2010 11:13 am Post subject:
Availability of the component is the percentage of time when system is operational. Availability of a hardware/software component can be obtained by the formula given below.
If we are using equipment which has ‘Mean Time Between Failure’ (MTBF) of 81.5 years and ‘Mean Time To Recovery’ (MTTR) of 1 hour:
MTBF in hours = 81.5*365*24=713940
Availability= MTBF/(MTBF+MDT) = 713940/713941 =99.999859%
Unavailability = 0.000141%
Outage due to equipment in hours per year
Unavailability =0.01235 hours per year.
Whereas, availability of an IS service or system (a combination of hardware/software components) is the percentage of time when the service is operational. Availability of an IS service can be obtained by the formula given below:
If an IS Service has an agreed uptime of 200 hours for a 30 day period (M-F 0800-1800hrs) and was not available for 3 hours during that time the availability of that IS Service is 98.52% for that period.
200/(200 +3)=0.9852 or 98.52%
However, when an IS Service is utilised by several different business units; and each unit has an individual SLA on place with IS Operations, the following formula can provide a means of calculating the end-user availability of a specific group, region or location, where the IS Service is delivered:
(Agreed Hours* No.of Users-Sum of (incident x Duration* % of users affected))/(Agreed Hours* No.of Users)*100
If an IS Service has an agreed uptime of 200 hours for a 30 day period (M-F 0800-1800hrs) and has a user community of 1000 users in four locations (Each location has 250 users) and one location experiences a single outage of 3 hours. Therefore the end-user availability of that IS Service is 99.625% for that period.
(200 *1000-(3 x 25%) )/(200*1000)*100=99.625 or 99.635%
Availability is typically specified in nines notation. For example ‘3-nines’ availability corresponds to 99.9% availability. A ‘5-nines’ availability corresponds to 99.999% availability.
Availability will be shown in service terms, as opposed to component, system or server terms as the business does not see the component level. All they know is that the email service is unavailable for example. This change in perception of unavailability will help drive IS to be more customer focused.
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum