SBS 2003 migration to 2011 suggestions


I will migrate an old SBS 2003 to new 2011 next week, so I ask for some suggestions. There are some specifics, so I want to make it smooth.

First of all, old SBS 2003 is used as plain simple FILE SHARE server, where even users are NOT joined into domain, but rather plain WORKGROUP users. The company admin did not have knowledge, so he went by entering each user's password on workstation and on server at the same time, so he established access to shared files.

There are also file shares, with permissions, fit for WORKGROUP users.

Beside that, users are using one Firebird SQL database on server via ODBC connector.

And that's all on old SBS 2003.
No exchange, no SharePoint, no WSUS...nothing.

So, what's your suggestion to approach to migration, so that users won't need to change anything?
I must say I have 2 days time window and there are 40 users.

My plan:
- run new SBS 2011 in migration mode
- migrate users and permissions to new SBS 2011
- leave users in WORKGROUP as they are so to avoid complications
- copy over SHARED files and folders, with permissions (for example, using WinRAR and stored NTFS permissions)
- install Firebird SQL on new SBS 2011 and copy over Firebird data
- check all FSMO are migrated to new SBS
- then decomission old SBS 2003
- wait for changes to propagate
- then change LAN IP address of new SBS 2011 to fit old IP address
- and rename new SBS 2011 to fit old name in order to avoid problems on users side

After this "kinda" migration customer will decide, if they will order also 2nd part, which would be migrating users from WORKGROUP to DOMAIN.

What do you think of my plan?
LVL 18
Andrej PirmanAsked:
Who is Participating?

Improve company productivity with a Business Account.Sign Up

Cliff GaliherConnect With a Mentor Commented:
That plan, honestly, is a recipe for disaster. First, you cannot rename an SBS server after install. Second, of you aren't going to use exchange, don't use SBS. Third, if you are going to keep people in a workgroup, don't use a domain don't use SBS.

Either commit to moving them to a proper domain, or don't. Remember the scene in Karate Kid where the old man talks about the right side and the left side of the road being safe, but then middle will get you squished just like a grape? Don't continue and propagate your predecessors mistakes. If you want a workgroup, fine. If you want a domain, fine. But you need to decide and commit.
I'm not fishing for points here
I would just like like to reinforce the comments above (should you have any doubt) that your proposed plan for migration will not work - do not do it!
You have clearly recognised that you have inherited a "problem" system
Presumably your objective is to replace the old server before encountering hardware failure ?
As above if you have no intentions of using exchange sharepoint etc why bother with SBS at all ?

The left side of the road is to replace the old server with new hardware on a workgroup basis - this retains the status quo

The right side of the road is to replace the old server with new hardware on an SBS basis but creating fresh domain and a clean install
Your problem will be with the timescale
But by doing a fresh install you will be able to setup (and test ) your new setup before your two day window (a weekend presumably?)

I have done seven SBS migrations and two clean installs in the past year
With 40 users and a clear preparation plan
You should be able to migrate 40 users in that time ( may be a long weekend though)
Andrej PirmanAuthor Commented:
Yes, you are both right - it is not the proper way, but rather a huge improvisation.

First of all - this company has been working in WORKGROUP environment for 8 years or so, and they are used to it. It is an 7:00 AM to 6:00 PM non-stop busy trade and they did not hire me to present them a plan, full of problems, with dozens of unknown variables, with some "domain" which they are afraid of...

So my idea was, as Cpmcomputers says, to assure new hardware base for the future. This would avoid hardware failure.
Deciding whether to try to convince them to migrate to DOMAIN I went to their office and examined few computers. was, how to say, an organized mess of everything. Permissions messed up, files and printers shared all across inbetween computers and old server, core business application running on server, but many helper applications also on workstations, shared across the workgroup.

I could make a plan of migration, but there is really NO WAY to assure, that if we start migrating in FRIDAY evening, they will all be full operational in MONDAY morning. Too many variables, too much mess.

So I am still into braking some rules to ensure their operability.

I'd go step by step, first new hardware and new SBS, then:
- as suggested, renaming SBS 2011 to match old server name, would save me a lot of work
- after a month or so, I'd start migrating clients one by one into domain
- Exchange is a plan, definitelly, but only after all clients would be migrated into domain

So, again, need opinions:
- renaming SBS 2011 is not an option anymore. I was hoping MS added some wizards for such procedure, but it is not the case, so I'll stick just to adding CNAME (or A) record in DNS to resolve old server name to the new IP of new server
- but keeping them all in WORKGROUP...well, I think it is the only option

Just some questions more:
- I have few days time left util next weekend migration, so should I go with FRESH install?
- ...or better with install in MIGRATION mode to easily move over FSMO and permissions?

Simplify Active Directory Administration

Administration of Active Directory does not have to be hard.  Too often what should be a simple task is made more difficult than it needs to be.The solution?  Hyena from SystemTools Software.  With ease-of-use as well as powerful importing and bulk updating capabilities.

Cliff GaliherCommented:
Imagine going to the doctor for a checkup because you've been getting headaches. The doctor comes back a few days later and says there are indications of a serious health issie.mNothing you've noticed yet except the headache,  but if you don't get on top of it, you might well collapse from a heart attack any day.

Do you really tell your doctor "oh, well,I'm really busy at work, can't you just give me a prescription for the pain. We can deal with that other stuff later." More importantly, do you really think your doctor would go along with that plan?? Of course not.

If you consider yourself an I.T. Professional, then take that work "professional" to heart. Only HALF of your job is providing the client with what they want. The rest is providing them with what they need, educating them on why they need it, and avoiding the pitfalls and not doing poor practices and things they DON'T need. This is not trivial and is often daunting.

But your reputation is on the line here. This company has employees. They talk. They will make recommendations to other friends that have small businesses. If at some Christmas party they start griping about their bosses, the economy, and other things, a bad tech experience can be surprisingly high in the gripe-fest. Tey will compare notes, and you will pay the price. I strongly recommend the books "The referral engine" and "duct tape marketing" to any solo I.T. Pro. Because every solution you sell is also a marketing tool, NOT just a solution to that customer. Which means every incomplete solution is negative marketing.

Sit own, have a conversation with the client, a REAL one. Tell them that a domain is a bad idea. Educate them on why. Else,win why you need to get them on a domain, and get them on board with the time and cost to do so.

Andnifnthey go with option B, don't sell them SBS. Period. Full stop. Even in a workgroup, each user needs SBS CALs. SBS is more expensive than a standard server, the CALs are more expensive, the labor of moving clients "a few at a time" is more expensive because of the split permissions model, and you sure as HELL should charge for every bit of that labor....their bad decision should not turn into your charity work....which means in every way selling them SBS is bad for them, financially, technologically, and more.

It'd be far better to sell them a standard server license and standard CALs. Then when they are ready and finally on board with doing a domain, sell them SBS. Sure, they eat the cost of the standard server license. But that is only a few hours of labor which they'll spend anyways on trying to wedge SBS into a workgroup, so it is a NO LOSS financial proposition in the end.

In summary, if you haven't figured out yet, I cannot recommend going forward with any variant of your plan under any circumstances. Come up with a new plan, please please please. I have not only been working with SBS in depth for 12 years, but I have been mentoring I.T Pros and MSPs and helping them improve their business models and practices for 15, and was mentored myself by some industry greats before that. Us is as much business advice as it is technical. Please, I hope you need it.
cpmcomputersConnect With a Mentor Commented:
Have you committed to new hardware or software yet?

If you were to go the SBS route
Would this be your first potential migration

Is it likely that you may need a dedicated separate server for the line of business applications

I feel that you need to take a step back with this and consider what is best for both you and your client
(Seems like you feel pressured into taking "something" forward without sufficient definition of the business objective?)

I have 35 clients on SBS and it is a fine product in the appropriate business model
But it is not a panacea for all ills
Andrej PirmanAuthor Commented:
Again, all you ara correct, but still there are FACTS which I must take into account:
- the hardware (new server) and software (SBS 2011 + 35 CALs) they already bought
- they allocated 500 EUR budget for migration with one freelancer
- and they've got weekend as the only time window

I can postpone for a week or so, and I DID deal for 1000+ EUR for migration (sorry, we are fighting for sallary each month) and that's it.
I cannot reset situation and start over, because it's out of my control.
It is not my decision, but a compromise to get as close to final destination as possible within the budget and given facts.

Which means, I MUST do something with what I have available.

Even if I would be able to talk them into consideration to go with proper migration to DOMAIN in one step, it is very unlikely that ALL would work in Monday morning, which means, that it would cost more (= I am the expensive guy) and they will have post-migration problems a week or two after the migration (= I am bad guy, because their system does not work), resulting in their most possible decision to hire that amateur freelancer again (= no money for my salary).

It's easy to chose proper path when you are IT admin in enterprise. But fighting for ANY job in the collapsed economy, where each and every EUR is hard to spend for customer and even harder to earn by us....well, it's an all-compromise job, from day to day.
I am sys admin, but I repair icons on desktop, I deliver printer toners to customers and beside installing IIS, DNS, DC and other roles, I also dance around with vacuum cleaner, sucking dust from 10+ years old computers.

Could you reconsider your suggestions? :)
Seems like they want you to turn their Skoda into a Ferrari and pay you nothing for your trouble :-)

If you have any option at all to walk away from this situation - I am sure you would take it?

To be totally honest I have not done an SBS migration where the users are in a workgroup, so would be guessing which option (Migration v Clean Install) would be "best".
You say they have an "Amateur" freelancer for 500 Euro's Has he done anything yet or is he available to work with you over a weekend? (is he capable and you can work with him)

Have you done a migration before ?
Andrej PirmanAuthor Commented:
Yes, I have done dozens of migrations before and I am, to be honest, in best case 70% familiar with by-the-book proper migration scenario. Mostly due to bad or insufficient preparation, which is crucial for smooth migration.

Maybe I was not clear enough about WORKGROUP organisation.

User are all in old SBS 2003 properly created in AD, but none of client computers is joined to domain, but rather left in Workgroup.
To access SBS 2003 shared folders and to match permissions, usernames and passwords are then created on each workstation to exactly fit SBS 2003 AD users/passwords. So they were able to share folders on SBS 2003

So, my thinking...
If I go with MIGRATION install mode with new SBS 2011, at least I would preserve SID's so username/password combinations for existing users would be preserved.
If I then copy over SHARED folders and files, I would get a copy of functional OLD environment on new hardware.
And that would be ALL for first step (within budget limits) - just a migration to newer hardware (and software).

Only after that, in next months, I would see if they will sign a contract with me for 1 year IT Service, and in that case we'd proceed with real migration, sector by sector:
- name/surname users would be slowly left over on behalf of new organisation, based on role/sector model (because employee fluctuation is huge there, mostly students)
- shared folders, printers and applications on workstations would be organised properly, applications moved to server, shares too, like it should be in serious company
- etc...

Meaning, I do not plan to do some lousy migration once and left it over, but rather first to assure thair hardware would not fail in next year, for some decent price for the first kick. So they should be happy and sign Service contract with me.
Only then the "real" migration would begin.

The freelancer I mentioned was kicked off due to my explanation on how things should work, so they will NOT hire him again...except if I would set my price too high.
compromise, I say :)

BTW... In order NOT to migrate old "shit" from old SBS 2003, for example, Exchange, SharePoint and WSUS settings to new SBS 2011...would it be a good idea to simply shut down those mentioned services on OLD server prior to migration?
Or would new SBS 2011 in MIGRATION mode pick all those settings from old one anyways?
Cliff GaliherCommented:
First off, let me say that I think you are setting yourself up for failure. Yes, I know the economy is depressed. But all the *more* reason to be careful. A short-term paycheck is *not* worth the long-term damage it does to your reputation. So many consultants live contract to contract exactly because they accept every job that comes along.

You have a simple choice. You can try to be the low-cost guy. In American terms, the low-cost big box store around here is Wal-Mart. It has a reputation for being cheap, but it is also mocked for the types of customers it attracts. They make money by sheer volume.

So if you are content to work your fingers to the bone then sure, you can be the Wal-Mart of consultants. You'll make money by doing work in volume. 80+ hours a week. Personally.

Or you can be the Lexus. Mercedes Benz. Cadillac. The higher end model that costs more but that the person paying the premium gets things that you can't get at Wal-Mart.

So this client bought a server, bought SBS, bought CALs, all without consulting you, and then tried to low-ball you on the migration fee? If they are willing to do that now, what is to stop them from buying some eBay crap printer in a month and then expect you to install it on all of their machines. For free. What will prevent them from deciding *every* problem they have for the next two months is migration related and that you should fix it for the $500-$1000 they are paying you. For free. You can tell a lot from a client by how they treat you up front.

To put it in perspective, my rate is $150/hr. An SBS migration *starts* at 40 hours. Do the math. My *low* contract for a migration is $6,000. You may get twice as many migrations as me, but I'll still make 3x as much when it is all said and done. And I'll have worked less making it.

So you want an honest answer? No, I don't feel at all compelled to change my suggestions.

I would *tell* the client that they need to give you more time (and more money) to move to a domain. Or, if they want to stick with a workgroup, they *need* to buy a Windows Standard license and CALs. When they are ready to move to a domain later, they can use the SBS license and CALs then and sit on them until that point. And if they refuse those options, I'd walk away from the client. It isn't worth the hassle, headache, risk of reputation, and most importantly setting clear expectations to put up with that.

Handling clients is sometimes like raising a kid. If you don't set boundaries, they will think they can keep getting away with murder and the only person it creates stress for is you. You need to be firm on these boundaries up front now, so that they don't abuse you in the future. And yes, that sometimes means threatening to walk away, and being willing to, and every once in awhile actually follow through on that threat.

Of course ultimately this is your decision. But you asked if I'd change my recommendations or answers, and as I said above, no I will not. And my reputation is important enough to me that I can't even in good conscience give advice in implementing a system I know to be wrong and bad for you and your customer. So even if I had advice to give, I would not.

Which also means this response will be the last one I provide on the matter. If someone else wants to help, perhaps they will, but I am not convinced and so I shall hope you take my advice, and otherwise step out of the way for you to do as you will.

Good luck,

Andrej PirmanAuthor Commented:

I appreciate your stand and I must say some 4 or 5 years in the past it would be exactly my words, as you said. I was tutti-frutty guy before, messing with washmachines, computers and car electrinics, working as maintenance guy in large enterprise. Learned a lot of many stuff there, and I thought, hey, I know many things, I can fix this and that...why not make some more money by doing this ant that exactly.
But as years past, I recon that what I did was only spending time with tiny problems, leaving my tiny "me" in kind of status quo - no gain in personality, no reputation raised, no big money, no career - just some dozens of customers, who wanted to pull more out of me because I was nice, friendly and I had no guts for big repairs to invoice big bills.

Then I recognised my failures and started saying NO to those tinny demands from tiny customers, spending more time for me, my life, career and self confidence. I moved away to a big city, sent over 30 resumes and picked up a really promising job.
Gained engineering title over years, done MS exams, Vmware, networking and such crap and after few years I was one of top 10 engineers for Fujitsu-Siemens enterprise computing in our country.
My role was on contract maintenance of big state institutions, practically I have had contracts with all Ministries, Tax and Law institutions in our country. Working 8-12 hours a day I can say I've got all you mention, reputation, excellent salary, own flat, new sportscar, luxury home...

Well, then recession came, my company bankrupted, state institutions cut down budget for outsourcing to 10%, hundreds of companies, also my customers, seized down. And so did my dream job.
I had few options then and I decided to go on my own, actually with some freinds from previous company, starting our own business. But after a good start, one of our biggest customers, one of top 10 in our country, went bankrupt and took along my whole income for 1 year! A month later 2 other big customers cut the contract due to lowering expenses.

That knocked me down on my knees, being left over with 10 small companies to maintenance by contract, which was just enough grasp enough $$$ for some minimal salary...but I had to pay my principals a sum of 1 year salary, which I could not. Could not get a credit on any bank, must pay taxes, insurance... what I could only do was to sell ALL what I had, TV, computers, my sportscar, my flat...
And, here I am now. With 4 open actions on Curt of the Law, blocked all bank accounts, being forced to partly do legal jobs and partly moonlighting to survive.

I have no option to be picky about customers and job I take.
If I have an option to do some GREAT job, I do probably better than many others. But If I get a chance to do ANY job, even some low-budget crap - what should I do? Should I decline it and then have no money for gas, phone and bills?
Or should I take even the worst crap, just to get thru?

So, I was actualy not asking whether should I proceed or not, but rather HOW to proceed in this situation. I WILL do it my way or a similar way, because I have no option to do it properly. So my query here was to get as much suggestions as possible to do minimal damage at start, so I would have maximum options to finish it later in a properly manner.

And again - I really appreciate your time taken to explain what you think is correct, and yes, you are right, but I am afraid not for me in my time and my space ;-)
Andrej PirmanAuthor Commented:
For those who are still interested into giving me some TECH advise on this crap of job, which I NEED to do this Friday, I opened another thread here:

I decided to split points for quality discussion and reward time you took to debate with me here. Separate points will go for tech advice in above mentioned 2nd thread.
I am sorry for your trouble
Cliffe makes valid points (but perhaps underestimates quite how bad the economy is in many parts of Europe )

Reviewing your situation and your options
The one thing that does occur strongly to me is that Cliffe is quite right in that the client is very likely to blame every conceivable shortcoming in their network on your migration for months afterwards
The danger is you will not get paid at all?

Do please protect yourself with regard to your payment terms and don't leave if til the "end of the project" to get paid

I will monitor your new thread (and contribute ) if I feel able to assist you
Best wishes
Andrej PirmanAuthor Commented:
CPM, thanx for feedback!

As I told, I understand perfectly what Cliff is talking about, but I cannot afford to be picky at this time.

The company is actually quite healthy, but very strict about plans. With help of their previous caretaker (that freelancer I mentioned) they already made a plan for new hardware and software some months ago, and they allocated 2.000 EUR for new server. Yes, 2k!
Then their internal manager got suspicious about knowledge and expertise of this freelancer, and he somehow came to me for advice.

Of course, I had 3 hours meeting with him after I examined their network and despite of me being polite, they fired freelancer the very next day. They asked me to make a plan and to not overkill the planned budget too much.

I made a plan for 6.000 EUR, of which there is approx 5k for server and software, and 1k for 1st step of migration. It took 1 month of phone calls, meetings until they finally approved 3-times bigger budget.
Beside I offered them a 1-year contract, so to participate more closely and to do proper migration in next months.

So, as you recognised my situation quite well, I must compromise very much at beginning, but if I would be careful enough and if I do not make any crucial mistake at this first PARTIAL migration, I might be able to reach win/win situation in next months:
- hardware renewed
- software too
- users migrated properly
- from external MAIL moved to internal Exchange
- I will sell them SSL certificate + SMTP proxy (which we offer)
- and I will have constant monthly income from the new contract

All happy, all fine.

And yes, they are strict payees - they pay in average 1-7 days before due, so I am not afraid not to get paid :)
Andrej PirmanAuthor Commented:
Dear CPM and Cliff,

after reconsidering the situation and testing the manual migration process, I decided to go with FRESH install.
Since only user accounts were planned to migrate, and since they are in Workgroup anyways, it is much easier to manually type-in all 40 users & passwords on new server, copy SHARES & permissions with, for example, WinRAR, and we've got a copy of OLD Workgroup environment on new server.

Also, previous Wrokgroup environment was based on Static IP addresses, which were crucial because of some other applications in network. I decided to use DHCP reservations instead.

The above mentioned procedure was successfully finished without any problem on TEST users and test data, so I assume the remaining 38 users should also "migrate" with no problems.

Thank you for participating in my decision.
Good luck with this :-)
The left side of the road I think ?
It gets you to objective one of protecting from hardware failure
If and when your clients can make a further budget provision I would look to moving
The clients to dhcp (easier to manage)

And perhaps look to add new pcs/users to the new domain
Likewise retiring old users/pcs from the workgroup

I think all round you have made the correct decision

Do post again to let us know how it goes
Or if you need any specific (technical) help
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.