• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 525
  • Last Modified:

SLA's for Windows Servers/Guests

Can I ask a few basics on service level agreements when it comes to windows servers, may be geared towards more client side management or contracts management type posters, or on the flip side management on the service providers:  

1)      What kind of SLA’s above and beyond “up time/availability” are typically included when creating a new IT contract with a 3rd party service provider.

2)      And how can an auditor check for compliance with the agreed SLA’s, i.e. say the SLA says xyz windows file server needs to be up 95% of the time, how can you check it was? Are there tools in windows server, or is it usually a 3rd party auditing suite that will give accurate up/down times?

3) In windows based networks aside from application specific servers, which servers typically need a higher “up time” than others?

4) How does a company compute the SLA’s for a server, i.e. why do they not just ask for 100% up time for every server in the domain? Why give flexibility and scope for down time, does more up time = more cost? If not why not just ask for the best for each server?

5) Can you get a server 100% “up time”, or what kind of issues/actions require “down time”, can you give me some examples or what takes a server down for a period of time?
0
pma111
Asked:
pma111
  • 6
  • 3
2 Solutions
 
pma111Author Commented:
And do you have SLA's per server typically in a contract, i.e. would the high importance servers stand out a mile? Or is it not really a per server SLA document.
0
 
Nagendra Pratap SinghCommented:
Ans 1:

You have P1 servers, P2 servers and so on. P1 server would have an SLA of say 2 hours. P2 like 4 hours.

This is the SLA to fix, there is another SLA to respond. That is like 30 mins max.

Ans 2: If you have automatic ticket generation then SLA is easy to computer. All ticketing systems have it.

Ans 3: Email, fax server have P1 priority. E commerce portal has P1 priority too.


And 4: Near 100 % is possible. You need to use clustering and other stuff but 100.00 % is not possible.

Anyone promising that is lying.

Ans 5: You may not get 100% uptime from a server but there are ways to get 99%  service uptime. Outlook can work without an exchange server for some time. Previous versions will let you see last 2-3 copies without using a backup system.

VMware HA, Clustering, Load balancing, File replication, mailbox replication, log shipping etc can increase uptime by a lot.

Servers are taken down for patching, hardware repair/upgrade and configuration change.
0
 
Ayman BakrSenior ConsultantCommented:
1)      What kind of SLA’s above and beyond “up time/availability” are typically included when creating a new IT contract with a 3rd party service provider.


- Request Response Time - Request Resolution Time - Support hours availability - Escalation process - Maintenance Window day and downtime.

2)      And how can an auditor check for compliance with the agreed SLA’s, i.e. say the SLA says xyz windows file server needs to be up 95% of the time, how can you check it was? Are there tools in windows server, or is it usually a 3rd party auditing suite that will give accurate up/down times?


Usually the calculation is done through a formula taking into account downtime against uptime. See this link for more info: http://www.quantumofgeek.com/2010/02/calculating-server-uptime/

3) In windows based networks aside from application specific servers, which servers typically need a higher “up time” than others?

It depends! You might want to have all your servers at the same availability level provided you have the budget. However, you might focus on your critical core business servers. For example: backend servers holding your critical data, domain controllers (the main life of your domain of which without your whole domain will die) etc...

4) How does a company compute the SLA’s for a server, i.e. why do they not just ask for 100% up time for every server in the domain? Why give flexibility and scope for down time, does more up time = more cost? If not why not just ask for the best for each server?


It is impossible to achieve the 100% availability of servers. With all the technologies to ensure higher percentage of uptime you can reach the 99.999% limit. The higher availability implies more technologies, data center, cooling etc... are invested thus implies more cost. This is so because machine eventually get 'old' and are prone to many issues, like hardware failure for example. This means you need to maintain them (the more you maintain and care for your car, the more you improve its life expectancy) - in other words you need some downtime for the maintenance, also called maintenance window.

5) Can you get a server 100% “up time”, or what kind of issues/actions require “down time”, can you give me some examples or what takes a server down for a period of time?

- Hardware Failure and thus replacement of hardware - Maintenance Window etc...
0
Get free NFR key for Veeam Availability Suite 9.5

Veeam is happy to provide a free NFR license (1 year, 2 sockets) to all certified IT Pros. The license allows for the non-production use of Veeam Availability Suite v9.5 in your home lab, without any feature limitations. It works for both VMware and Hyper-V environments

 
pma111Author Commented:
Thank you. Can you help with "And do you have SLA's per server typically in a contract, i.e. would the high importance servers stand out a mile to a non techie? Or is it not really a per server SLA document. "

Also - can you give some idea of tasks that fall into "maintenance"....
0
 
pma111Author Commented:
Also, "- Request Response Time - Request Resolution Time - Support hours availability - Escalation process - Maintenance Window day and downtime."

Does performance fall into any of these? When you mention requests respose/resolution could you provide a few examples of "requests" just to aid my understanding.
0
 
Ayman BakrSenior ConsultantCommented:
Well I have not been involved into contracts so I wouldn't be able to tell you about SLA's per server or as a whole. But I believe that you can detail in the contract specific service agreement for specific servers.

Performance would be the measure of how stable and well is the server doing its intended functions. What I meant with requests is the tickets created by the client, when they face incidents or issues deviating from normal operation, and submitted to the service provider for their support team to troubleshoot and resolve. For example, let's say the email server hosted by the service provider is no longer sending emails. You as a client will create a request/ticket to the vendor to resolve the issue. The contract specifies the response time for this ticket - for example within 30 min from logging the request. It will include also the resolution time - for example max. 4 hours. You can also include penalty agreements in this contract for violating any of these timings for example.

Maintenance activities will include: patching, updating services/service packs/drivers, upgrading apps, restarting servers to free cache and memory etc...

For what it's worth SLA is the term used for agreements between entities of the same organization. Agreements with service providers or vendors are usually called Underpinning Contracts.
0
 
pma111Author Commented:
FAO - npsingh123 - what does the "P" represent in P1, P2 etc. What formula do you use to identify which servers fall into which category. Can you provide an example of the types of server in the P2 category.

All - Can SLA's be different per server. I.e. if you have an issue with a P1 system, may the SLA resolution time be faster than say for a desktop PC, or a P3 server?

What "document" will usually have a copy of the SLA's? And in what format will they be shown?
0
 
pma111Author Commented:
And do you "P" every asset in your infrastructure, or just "servers" ?
0
 
pma111Author Commented:
And what is your P list referred to as, so I can ask for similar here....
0
 
Ayman BakrSenior ConsultantCommented:
SLA would be part/section of the contract or In a document called the service catalogue.

For more on SLAs and what to include in it see this link:

http://en.wikipedia.org/wiki/Service-level_agreement
0

Featured Post

Automating Your MSP Business

The road to profitability.
Delivering superior services is key to ensuring customer satisfaction and the consequent long-term relationships that enable MSPs to lock in predictable, recurring revenue. What's the best way to deliver superior service? One word: automation.

  • 6
  • 3
Tackle projects and never again get stuck behind a technical roadblock.
Join Now