ITIL Availability Percentage and what the percentages mean : ITILNews.com

ITIL Availability Percentage - What the percentages really mean [7 comments]

Ultimately, we are in business to deliver a service and more often than not the effectiveness of the service is gauged by its overall availability, normally as a percentage. The availability of a service is determined in the Service Design stage and has an impact on how the service is developed. The 'target availability' is contained within the Service Level Agreement (SLA) and often provides a topic of considerable debate when Service Availability negotiations take place. If you have ever wondered what the percentage figures equate to in hours, minutes and seconds then take a look at the following table.

For the purposes of this exercise we are assuming the service in question is required 24/7/365.

Availability %	Downtime per year	Downtime per month*	Downtime per week
90%	36.5 days	72 hours	16.8 hours
95%	18.25 days	36 hours	8.4 hours
98%	7.30 days	14.4 hours	3.36 hours
99%	3.65 days	7.20 hours	1.68 hours
99.50%	1.83 days	3.60 hours	50.4 minutes
99.80%	17.52 hours / 1051.2 minutes	86.23 minutes	20.16 minutes
99.9% ("three nines")	8.76 hours / 525.6 minutes	43.2 minutes	10.1 minutes
99.95%	4.38 hours / 262.8 minutes	21.56 minutes	5.04 minutes
99.99% ("four nines")	52.6 minutes	4.32 minutes	1.01 minutes
99.999% ("five nines")	5.26 minutes	25.9 seconds	6.05 seconds
99.9999% ("six nines")	31.5 seconds	2.59 seconds	0.605 seconds

Perhaps the 'holy grail' of any IT service is the ability to achieve 'six nines'. I have been scratching my head but I still cannot name a service that achieves such an availability figure. If you know of such a service then please let us know by completing the 'Your Comments' area below.

7 VISITOR COMMENTS

2010-12-18 by "Chrispope"

Hi, thanks for this, very useful indeed and similar to an exercise I am currently going through. What isn't clear here is when should the service be available? Is it assumed to be 24x7 or Mon-Fri 7am to 7pm?

Interested to undersstand and then this puts into context the figures in the downtime per year column.

I am new to the site and have found some really useful information on here, crossing the theory vs practical experience boundary is not an easy one!

2011-02-14 by "gchiock"

The contracts must notify the range of the time where the availability will be up. Perhaps it should be 24 x 7 or other values. In those cases, you must calculate the availability values based in the number of hours considered.

Reply on 2011-02-25

Thanks for taking time out to make this point.

2011-04-28 by "alan.legal"

Is this calculation inclusive or exclusive of scheduled maintenance?

Reply on 2011-05-06

Great question ! The calculation does not include scheduled maintenance or agreed unavailability.

The type of service will normally dictate the level of availability required, for example, life supporting services such as fire, ambulance, police, aviation... such services may well be duplexed to enable availabiity to remain unaffected as scheduled maintenance takes place.

2011-05-16 by "cfrancum"

First, let me congratulate you on the article. Well explained.

Per Availability of 6 nines: I worked on a project for the FAA called Voice Controlled Switching System. It had an availability target of 7 nines - and achieved it. VSCS provided the Air Traffic Controller communications to airplanes. As it was explained to us: Do you want your ATC to lose communications when YOU are landing??VSCS may have been replaced by now.

2012-11-05 by "sarath.kolla"

I presume DoD's ThermoNuclear War Watch Programs are six 9's if anything in this world!!

Reply on 2012-11-09

Does anyone have any feedback on this question ?

2014-12-04 by "robert.vruink"

When looking at "Availability" consider for one moment how you might apply it to the international air transport industry and "Service Unavailability" of 6 nines+ measured per passenger mile.

2014-12-28 by "billh"

Active/active systems using HP NonStop servers typically receive this level of availability. There are several banks and telecoms using these systems that have not had an outage in decades (see Case Studies in www.availabilitydigest.com).

Active/active systems use two or more nodes that are actively processing transactions for the same application. Their application databases are kept in synchronism via bidirectional data replication. It a node fails, all that is required is to route further transactions to a surviving node. Failover can be achieved in seconds, and there are no failover faults.