Threat level definition
Search:
Home
Bugtraq
Vulnerabilities
Mailing Lists
Jobs
Tools
Vista
News
Infocus
Foundations
Microsoft
Unix
IDS
Incidents
Virus
Pen-Test
Firewalls
Columnists
Mailing Lists
Newsletters
Bugtraq
Focus on IDS
Focus on Linux
Focus on Microsoft
Forensics
Pen-test
Security Basics
Vuln Dev
Vulnerabilities
Jobs
Job Opportunities
Resumes
Job Seekers
Employers
Tools
RSS
News
Vulns
Security Research
BS 7799/ISO 17799
Back to list
|
Post reply
re: Maximum SLA?
Feb 28 2008 02:20PM
Robert MacDonald (Robert MacDonald Haworth com)
Let's try again. Sorry for the html.
----- Original Message -----
From: Robert MacDonald
To: Worrell, Brian <BWorrell (at) isdh.IN (dot) gov [email concealed]>; Kosala Atapattu <kosala.atapattu (at) gmail (dot) com [email concealed]>; Scott <whip (at) supportmenot (dot) com [email concealed]>
Cc: bs7799 (at) securityfocus (dot) com [email concealed] <bs7799 (at) securityfocus (dot) com [email concealed]>
Sent: Thu Feb 28 08:40:42 2008
Subject: re: Maximum SLA?
Wouldn't the total outage time depend on the wording of the SLA?
If the SLA excludes maintenance downtime (weekly/monthly tech windows),
then the ~5 minutes could be achieved much easier.
________________________________
From: Worrell, Brian
Sent: Thu 2/28/2008 7:39 AM
To: Kosala Atapattu; Scott
Cc: bs7799 (at) securityfocus (dot) com [email concealed]
Its been a while, but if I recall the number, at 5 9's, that means less
than 5 minutes downtime for the entire year. Being realistic, even in a
perfect environment, is it really possible to have zero issues that
would cause any down time?
Thinking from a clustered network, with clustered servers, etc. what's
the best case for failover on them, 20 seconds?
My thought would be that anything more than 5 9's would be hard to do,
over then entire system.
Just my two coppers....
(Wiki ref on uptime math http://en.wikipedia.org/wiki/Uptime )
-----Original Message-----
From: listbounce (at) securityfocus (dot) com [email concealed] [mailto:listbounce (at) securityfocus (dot) com [email concealed]]
On Behalf Of Kosala Atapattu
Sent: Thursday, February 28, 2008 6:16 AM
To: Scott
Cc: bs7799 (at) securityfocus (dot) com [email concealed]
A dynamic web page, like a portal service. Which involves Websphere
clustered infrastructure with a DB backend, on a HA clustering.
I just want to know whether there is anybody who managed to achieve 5
9's or more for a setup as above.
Even for a static web page, can you achieve more than 5 9's uptime, for
example 6 9's?
Kosala
On Thu, Feb 28, 2008 at 2:06 PM, Scott <whip (at) supportmenot (dot) com [email concealed]> wrote:
> It really depends on what service you are providing at 5 9's. It is
> relatively easy to publish a simple static web page to 5 9's uptime,
> but a full company IT infrastructure at 5 9's is a different kettle of fish.
>
> What exactly are you wanting to know?
>
> Cheers,
> Scott
>
> -----Original Message-----
> From: listbounce (at) securityfocus (dot) com [email concealed]
> [mailto:listbounce (at) securityfocus (dot) com [email concealed]] On Behalf Of Kosala Atapattu
> Sent: Thursday, 28 February 2008 6:22 PM
> To: bs7799 (at) securityfocus (dot) com [email concealed]
>
> Dear All,
>
> Is there anybody in this list who adhere to a SLA of more than
> 99.999% (or 5 9's). What is generally required to achieve a SLA more
> than 5 9's?
>
> --
> Kosala
>
> blog.kosala.net
> www.linux.lk/~kosala/
> http://www.kosala.net/
[ reply ]
Privacy Statement
Copyright 2008, SecurityFocus
----- Original Message -----
From: Robert MacDonald
To: Worrell, Brian <BWorrell (at) isdh.IN (dot) gov [email concealed]>; Kosala Atapattu <kosala.atapattu (at) gmail (dot) com [email concealed]>; Scott <whip (at) supportmenot (dot) com [email concealed]>
Cc: bs7799 (at) securityfocus (dot) com [email concealed] <bs7799 (at) securityfocus (dot) com [email concealed]>
Sent: Thu Feb 28 08:40:42 2008
Subject: re: Maximum SLA?
Wouldn't the total outage time depend on the wording of the SLA?
If the SLA excludes maintenance downtime (weekly/monthly tech windows),
then the ~5 minutes could be achieved much easier.
________________________________
From: Worrell, Brian
Sent: Thu 2/28/2008 7:39 AM
To: Kosala Atapattu; Scott
Cc: bs7799 (at) securityfocus (dot) com [email concealed]
Its been a while, but if I recall the number, at 5 9's, that means less
than 5 minutes downtime for the entire year. Being realistic, even in a
perfect environment, is it really possible to have zero issues that
would cause any down time?
Thinking from a clustered network, with clustered servers, etc. what's
the best case for failover on them, 20 seconds?
My thought would be that anything more than 5 9's would be hard to do,
over then entire system.
Just my two coppers....
(Wiki ref on uptime math http://en.wikipedia.org/wiki/Uptime )
-----Original Message-----
From: listbounce (at) securityfocus (dot) com [email concealed] [mailto:listbounce (at) securityfocus (dot) com [email concealed]]
On Behalf Of Kosala Atapattu
Sent: Thursday, February 28, 2008 6:16 AM
To: Scott
Cc: bs7799 (at) securityfocus (dot) com [email concealed]
A dynamic web page, like a portal service. Which involves Websphere
clustered infrastructure with a DB backend, on a HA clustering.
I just want to know whether there is anybody who managed to achieve 5
9's or more for a setup as above.
Even for a static web page, can you achieve more than 5 9's uptime, for
example 6 9's?
Kosala
On Thu, Feb 28, 2008 at 2:06 PM, Scott <whip (at) supportmenot (dot) com [email concealed]> wrote:
> It really depends on what service you are providing at 5 9's. It is
> relatively easy to publish a simple static web page to 5 9's uptime,
> but a full company IT infrastructure at 5 9's is a different kettle of fish.
>
> What exactly are you wanting to know?
>
> Cheers,
> Scott
>
> -----Original Message-----
> From: listbounce (at) securityfocus (dot) com [email concealed]
> [mailto:listbounce (at) securityfocus (dot) com [email concealed]] On Behalf Of Kosala Atapattu
> Sent: Thursday, 28 February 2008 6:22 PM
> To: bs7799 (at) securityfocus (dot) com [email concealed]
>
> Dear All,
>
> Is there anybody in this list who adhere to a SLA of more than
> 99.999% (or 5 9's). What is generally required to achieve a SLA more
> than 5 9's?
>
> --
> Kosala
>
> blog.kosala.net
> www.linux.lk/~kosala/
> http://www.kosala.net/
[ reply ]