Rarely do I completely delete a posting but I am going to in this case. If you have a torn state C34 install, we have a supported workaround.
Please use the new information for this located here: http://blogs.technet.com/b/joscon/archive/2011/05/10/supported-workaround-for-torn-state-installations-on-windows-7-sp1.aspx
I have never seen Microsoft so thoroughly side step its obligation to fix an issue but they have. Microsoft not fully testing the WSUS rollout of SP1 caused the failed SP1 installs. Your own support services told techs and end users to use the pending.xml "fix". Now after what 4 weeks you’re telling us that Microsoft is not willing to accept the blame by either providing a solution or sending checks to those end users who will now be paying me to fix your mess.
Way to go Microsoft, time to turn off updates! You have completely destroyed any trust that was remaining in patch management with this latest and I key on that word, LATEST failed patch.
Agreed. We have also turned off all updates from WSUS.
I don't want more destroyed machines with no fix from Microsoft.
"Rebuild the machines" is not an acceptable "fix".
I agree as well. This is unacceptable. We have over 70 computers with this problem and we are supposed to reimage all of them because of something Microsoft screwed up? Right. We have also turned off WSUS updates.
"But at this time rebuilding the OS is the only Microsoft supported solution" were you serious when you wrote this? Rebuild 75 clients? save data on 75 clients and put it back? r u coming to do all this on my 30 customer each having more than 50 clients?
I also agree that this is an unacceptable resolution. So can we send the bill to MS for the time it takes us to rebuild all of these machines? This is another reason to not trust the "security focused" (MS words, not mine) Microsoft solutions. Thanks once again for another lovely Vista like experience..
How am going to get my customer to trust Microsoft products in the future?
First, I am sorry that you are the "fall guy" for Microsoft on this one (at least public). I and others do appreciate these updates.
I have to say, I find this extremely dissapointing and lacking from Microsoft. I have many sites that had this issue, and my small company is the one to bare the cost of repairing our customers systems while the big MS sit happy with their licence sales.
I tell you what. How about you send me the email of the person in Microsoft that I can send the bill to for all this work that the WSUS version of this update has caused? Then you dont need to convince me that a version of the installer isn't needed to fix this.
Microsoft should release an inplace upgrade build that people in this state can use.
Thanks for listening (again)
Thanks. Rest assured that these comments are being seen. I know in the end that might not mean much but it's all I have right now.
I'm sad. Disappointed. I keep feeling like these folks beta tested SP1 and didn't get even a thank you or a swag gift. And if these were SBS boxes themselves that got nailed with c34's where you cannot trivially reinstall these boxes and hopefully have a backup I would not be so accepting of this position here.
In the Microsoft partner forums even support personnel pointed to the pending hack. So even Microsoft's own support staff were pointing to workarounds that ultimately ended up breaking boxes.
Microsoft changed how Win7 was deployed and didn't communicate to their customers. For that reason alone, I still feel that Microsoft should have done more than just blow this off like it feels happened here.
You say that the product group understands -- no I really do not think they understand how deeply this is those impacted not trust Microsoft going forward. Trust in patching was broken. Trust in servicing Win7 sp1 was broken.
My trust, my hope that Microsoft would do the right thing by these customers has been broken.
I think I'm hurt most about that. I really thought the right thing would be done here because I still feel the first failure was the decision early on to not flag this service pack as exclusive. That decison turned Microsoft customers into testers of this service pack without their approval.
So no, I don't think the product group understands. This has a human impact I don't think they are factoring in.
Again, like Nick, I want to deeply thank Joseph for being the one being placed in this position. It's not fair to him at all. I'm sorry that you were placed in this position and thank you for your work here.
Bottom line, I'm saddened this occurred.
While I go to dozens of clients with dozens of machines to fix these issues for free, who is going to make my mortgage payment? I certainly can't be making money while I'm fixing these for free. Just like BP's response to the gulf disaster, I think Microsoft needs to respond with a similar program as the fixing of this issue essentially puts me and my clients temporarily out of business unless you can come up with a better solution. The handling of this has been poor and the solution is no solution at all.
Will running an in-place upgrade do what's required to get the machines out of a torn state or do we have to wipe and install? I know rebuilding about 30 machines isn't a monumental task for many admins that read this blog, but it's a considerable chunk of time for just myself and an assistant at our company, staggered over weeks and coordinating schedules with users to get their machines and keep them for the needed time.
Also, thanks for all of your communication joscon. Obviously, this is a frustrating situation for everyone, yourself included, but we'd be worse off without your posts and answers. I hope those watching at Microsoft will acknowledge the attention you have paid and continue to pay to this issue.
An inplace should suffice as its really just rebuilding the Windows directory. You could then use something like USMT to do a hardlink migration (USMT hardlinks that is) and move the user data back to the old directories. It's covered here: technet.microsoft.com/.../dd883247(WS.10).aspx
And, you're welcome RV. I'm glad that this is providing value to some of you even though I know this isnt the outcome to this that all of you were hoping for.
So what are you people going to do about this? Still wait for fix? Try the earlier posted workaround by Kelvin Aston or just giveup and reinstall OS? I really want this fixed and i simply cannot reinstall all the systems i have in torn state, period.
Feels like i am refreshing this blog too often and hoping for solution.
The currently supported solution is posted in the blog and thats a rebuild.
How long have y'all been doing IT?
THIS KINDA CRAP HAPPENS!
Hate it, but it does. If OS's and programs did everything they are supposed to do.. a lot of us would be out of jobs and Microsoft would probably rule the world. There has been only one perfect OS.. wait sorry that was person. There has never been a perfect OS ( though like everything else in the world someone will state otherwise ).
Now to tech stuff.
I need clarification on what exactly determines if a system is torn or not. What are the requirements to be torn and thus be reloaded? I originally tried running just the script, that produced 0 nodes changed and still the fatal error. I went in and deleted a checkpoint 0000000000000000.cdf-ms entry, still got the error. Ran a script from either Joseph or a user on one of these pages ( I looked at so many I honestly don't know and don't feel like investigating at this point ), that produced 2 node changes and a running Windows 2008 server. Because I made any changes to pending.xml is that an automatic "TORN" server?