should we wait for server 2012

[gibbo_ap]

we are putting in a new server in a primary sometime this year, and looking at the dates for server 2008r2 is it worth the wait for 2012? i know extended support is until 2018 but we arnt long off 2012.

 

the sensible part of me thinks we should wait until 2012 is at least a year old.

[SYNACK]

I'd be inclined to say you should just have an EES agreement so that you can just install the most current one whenever you want to.

[Rydra]

I'd be inclined to say you should just have an EES agreement so that you can just install the most current one whenever you want to.

 

Rarely cost effective in Primaries. By the time you NEED the upgrade, the kit it's running on is out of date anyway.

 

I would hold off on S2012, firstly to get the bugs out of it (do you really want your network running on a partially tested buggy system?), and until you're client machines are at a state where they NEED the upgrade to use all the features.

 

Since most schools have barely left XP for Windows 7 it seems silly to upgrade to a newer OS to support kit that isn't used yet.

[sparkeh]

Go with 2008r2 - its well tested, patched and stable.

Early adoption is rarely a good option IMO.

 

Though I would also look at EES

[SYNACK]

Rarely cost effective in Primaries. By the time you NEED the upgrade, the kit it's running on is out of date anyway.

 

I would hold off on S2012, firstly to get the bugs out of it (do you really want your network running on a partially tested buggy system?), and until you're client machines are at a state where they NEED the upgrade to use all the features.

 

Since most schools have barely left XP for Windows 7 it seems silly to upgrade to a newer OS to support kit that isn't used yet.

 

 

This is a lot of the problem that I think some schools have, they wait, not untill the need the upgrade or even need the upgrade but when the NEED it as there is no other option other than going back to chalk slates. If they stayed up to dateish upgrading incrementally when it is right to there would be less hassle overall with compatibility, training and support.

 

If people were following the same patterns even back when XP came out they would have gone from Windows 3.0 to Windows XP. Why keep making the same decisions when theyhave proved to make things much harder with regard to training, compatibility and flexibility.

[mthomas08]

Rarely cost effective in Primaries. By the time you NEED the upgrade, the kit it's running on is out of date anyway.

 

I would hold off on S2012, firstly to get the bugs out of it (do you really want your network running on a partially tested buggy system?), and until you're client machines are at a state where they NEED the upgrade to use all the features.

 

Since most schools have barely left XP for Windows 7 it seems silly to upgrade to a newer OS to support kit that isn't used yet.

 

+1 on this one, we avoided Vista for this very reason and now look - thats almost forgotten compared to Windows 7. 2008R2 has worked great here *touch wood* .

[googlemad]

To be fair the upgrade path from 2008 R2 to 2012 will be a lot easier too, when we had to do 2003 to 2008 R2 it meant just building a lot of servers from scratch and migrating the actual data (SQL etc) over afterwards. Since it was a 32 bit to 64 bit upgrade path.

[elsiegee40]

Rarely cost effective in Primaries. By the time you NEED the upgrade, the kit it's running on is out of date anyway.

 

I would hold off on S2012, firstly to get the bugs out of it (do you really want your network running on a partially tested buggy system?), and until you're client machines are at a state where they NEED the upgrade to use all the features.

 

Since most schools have barely left XP for Windows 7 it seems silly to upgrade to a newer OS to support kit that isn't used yet.

Another +1 for this

[gibbo_ap]

we are going with ees with the school(s) but without the server os strap on, i have since reread the doc i was looking at and the support is not 2018 its 24months after the next sp

 

Microsoft Product Lifecycle Search

[Rydra]

Where I am I've started working on upgrading all hardware/software to the point where it can be better managed/controlled.

 

Hardware is 3 years for an upgrade (EG more memory, bigger HD where applicable), and 5 years for replacement. This is more because you don't want to rely on 5 year old kit, it's not really designed to last that long with really heavy usage.

 

OS's I would say don't touch new versions for a minimum of 12-18 months, and wouldn't consider a mass upgrade/rollout within 2 years of release. OS's should ideally not be used for more than 5 years after release, the exception being with XP, where vista was such a pile of fail it had to limp on for longer.

 

Other software (EG office suites) I would suggest a 5 year program, as it takes 1 year to iron out the bugs, and 2 more years for the 'new version' to become standard.

 

The servers are straight forward; 5 year replacement program, and stagger it by 2-3 years; so at no point do you have ALL new kit, you have AT LEAST one reliable bit of kit. This reduces the chances of mass failure due to hardware problems, or mistakes in config. Server OS's need to move with the prevailing clients. When the majority of your kit is not supported by the existing Server OS, it's time to eye up those upgrades.

 

EES from memory costs about 15-20% more than the outright purchase of the OS/CAL's, and is normally about a 3 year program, and is only cost effective if you keep upgrading every 3 years.

 

I am likely to upgrade to windows 7 by next summer, and will upgrade my main server to 2k8r2 this summer, giving me 3-5 years more on that OS and server 2008r2.