0

Ok so monday i go into work with a serrver with a bad hard drive on my raid 10 array. I ordered a drive and put it in wednesday (today) light are green and blinking meaning it is rebuilding.

Suddenly about an hour in while the server is running it goes down. Os error causing bsod. Restart and windows wont start. So i start diagnosing it. It still wont start but i think it is because the other drive now shows a red light.

Now i think hmmm im screwed i have to do a reinstall. So i panic and get my secondary controller up by using ntdsutil to get control. I finally get everything up and the original comes back i dont know how i dont know why it just did. The original still has the drive in bay one blinking orange and both drives blinking green.

So i go and test the network and i have net and remote desktop to all my servers but none of my clients have access to the full share.

In windows explorer i type \server\share i get one folder and thats it. Maybe one folder under that has full access the rest have a gray x on them and dont show anything. So i pull all control back to original dc because its working and now same issue.

Dhcp and dns werent right on the clients so i get it back to what they are supposed to be and it still does the same thing.

Any ideas or hell questions to help point me in the right direction are greatly appreciated. I have been working on this for 16 straight hours and i just cant give up because these shares are needed for our erp.

Improve this question
4
  • Holy stream of consciousness batman. Care to organize and/or insert at least a line break or three?
    – EEAA
    Nov 5, 2015 at 3:39
  • I am terribly sorry. I have been here since 630am and I really need help. I will edit when I get chance.
    – Matthew Spotts
    Nov 5, 2015 at 3:44
  • What do you mean by "active directory cant see shares"? Are you referring to SYSVOL? Or did you have some other custom file shares on your domain controller?
    – Ryan Ries
    Nov 5, 2015 at 4:43
  • You don't seem to have made any progress after 16 straight hours of working this, and from what I can see, you haven't even asked the right questions, and don't even know where to start. Hire someone to do what you can't. Lots of consultants would be thrilled to bill a dozen hours at a couple hundred bucks an hour to fix this for you.
    – HopelessN00b
    Nov 5, 2015 at 18:00

3 Answers 3

Reset to default
2

So it sounds like you seized the FSMO roles to another DC while the DC in question was down and then the DC that was down came back up? If so, you need to shut it down, purge it from AD, rebuild it and re-promote it. You can't reconnect it to the domain after you've forcibly seized the FSMO roles from it.

Improve this answer
10
  • I also re-seized the roles back
    – Matthew Spotts
    Nov 5, 2015 at 4:08
  • 1
    Honestly, I've never heard of anyone doing that but I'm pretty sure you screwed the pooch by doing that. The MS guidance is pretty clear and pretty stern in that a DC that has had the FSMO roles seized should never be reconnected to the domain.
    – joeqwerty
    Nov 5, 2015 at 4:15
  • 4
    I really don't know at this point. I'm assuming this is in an Enterprise environment and that there are users and a business that rely on this AD? If so, the best thing I can tell you at this point is to open a support case with Microsoft ASAP and let them guide you.
    – joeqwerty
    Nov 5, 2015 at 4:22
  • 2
    Do yourself a favor and run dcdiag /e on one of the Domain Controllers and then search the resulting log file for all instances of the word Failed and see what you find.
    – joeqwerty
    Nov 5, 2015 at 4:37
  • 1
    The only visible fallout is the shares. A flailing active directory doesn't always show visible signs, and it rarely shows serious issues right away unless you're looking for them. Please follow joeqwerty's suggestion - I haven't seen or heard of this happening either so I'd hate to start suggesting things to try if you should instead be making backups and recovering what you can.
    – Neil
    Nov 5, 2015 at 5:32
1

Ok So I found the solution joe pointed me in some direction and the rest I discovered by trial and error so I am going to do a bullet of the steps that led upto the problem and the solutiom.

  • PDC has critical OS failure during raid rebuild of the drive. This rendered the server useless.
  • Used ntdsutil to seize all the roles of the domain on the SDC
  • SDC never functioned as PDC althoughit appeared to be configured correctly
  • PDC array rebuild finished and recovered a backup to make the PDC a function server again but not as PDC because it had its roles stripped.
  • use ntdsutil to seize all roles of the domain to the PDC (this is about the time I posted on here)
  • PDC not operational on the domain and in fact cannot see the domain
  • a ping of the domain from the PDC shows it is pointed at the SDC which is no longer in control and was never functioning as PDC.
  • investigated the DNS server and pointed it back at the PDC
  • Used dcdiag /e and found more error than passes but one specifically caught my eye the error read to the affect that there were no GC servers all GC servers are down.
  • after more research I discovered this article. The Sysvol and Netlogon Shares Are Missing After You Restore a Domain Controller from Backup
  • Domain sprang back to life with in minutes

Now I believe all of the errors are slowly disappearing from dcdiag report

Please help others and contribute to my answer where I went wrong or where things could have been done better. I hope that my struggles will help everyone here.

On a side note I was on the phone with microsoft as Joe suggested about to give them credit card information when it dawned on me to ping my domain. I know most of you probably would have done that out of habit but the thought never crossed my mind until then.

Thank you all so much I know my first post and replies were probably the worst ever on here but I was tired and panicked. You guys are the best. Enjoy the rest of you day.

Thanks Matt

Improve this answer
0

If everything else is working, it seems like the thing to do at this point is make sure your array is back to optimal and then restore the shares from backup. You do take and test your backups, right?

Improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .