Ris v Normal Build

[garethEds]

Here we use RIS to build machines. It has been successful.

 

Some F series RM machines refuse to build using RIS due to an issue with the BIOS - so we hand build them and then update the BIOS.

 

Anyway - I've just taken a call from the technician telling me that hand built machines perform a lot better than RIS built machines. I'm quite happy with RIS - it works.

 

Is this the case or shall I send him around the school to rebuild every machine manually so we can see the performance increase he claims there is :-)

 

GJE

[Quackers]

I fail to see how a handbuilt is any faster than a RIS built, same files... only difference is you have a config file auto answering questions for you and the files come down a network cable instead of off a CD....

[garethEds]

I fail to see how a handbuilt is any faster than a RIS built, same files... only difference is you have a config file auto answering questions for you and the files come down a network cable instead of off a CD....

 

The advice given is probably politically motivated as he doesn't like the other technician who set our RIS up for us.

 

GJE

[PeterW]

Anyway - I've just taken a call from the technician telling me that hand built machines perform a lot better than RIS built machines. I'm quite happy with RIS - it works.

 

Souds like tosh to me

[flyinghaggis]

Different drivers possibly? Either that or something you've built into the RIS image that shouldn't be there when you captured it maybe?

[powdarrmonkey]

Somebody has a grudge...

 

I'd go with your plan of getting a comparison done. Imaging every machine in the school twice will soon shush him up.

[Guest salan]

Myself, I would say that IF the RIS is set up FULLY correctly for the machine, then its just as good.

The only time I could see that it woldn't be is if in the setup of the answer files etc there has been compromises made to make it more 'universal'.

[DrPerceptron]

I've suspected that handbuilt machines are quicker than machines we've deployed via WDS, I don't know what would make that so, but I've always suspected it.

 

Mainly because a reference PC I had was 10-15 seconds quicker logging on a new user and reacting to bigger, meatier applications.

[Michael]

It simply isn't true, however I would say great care and thorough testing is needed before deploying any image, using whichever method.

 

RIS is simply a deployment tool/mechanism whereas the image itself is what users actually use

[jsnetman]

Mainly because a reference PC I had was 10-15 seconds quicker logging on a new user and reacting to bigger, meatier applications

 

Probably because the comparison was done on the manually built machine before it was joined to the domain and its policies. If RIS is setup properly it automatically joins the domain and receives its policies therefore slowing the logon time.

[pagelad]

Cant comment on RIS but i just quickly deployed a WDS image and installed a machine to compare, both booted up at the same time

[dalsoth]

Can i also say that the F series work much better with a full case off and clean inside with toothbrush Get your techie on to that too ASAP. Dust waits for no man!!!!

[pete]

Like others have said, unless your RIS setup is broken, there shouldn't be a real difference.

 

You could (as a test) build one from scratch with slipstreamed XP cd, install the apps and then capture using WDS, and build another using the default image in RIS. Deploy both to identical machine and see if there's a difference.