none
Update kb4057903 fails in windows server 2012 r2 Hyper-V

    Question

  • While trying to install the update KB4057903 - Hyper-V integration components, the server freezes after restart on the screen "Getting Windows Ready" for about 15minutes. After that the server gets almost unresponsive.

    Managed to restart the vm via safe mode and stopped the update, but how can i install the update? What causes the problem? I had no problem with the other updates of January 2018. Error Details: 800F9023. In HYPER-V VM enviroment.

    It is also the exchange server and i have stopped all Microsoft Exchange Services.

    Sunday, January 14, 2018 12:48 PM

All replies

  • Might also try from a clean boot.

    https://support.microsoft.com/en-us/help/929135/how-to-perform-a-clean-boot-in-windows

     

     



    Regards, Dave Patrick ....
    Microsoft Certified Professional
    Microsoft MVP [Windows Server] Datacenter Management

    Disclaimer: This posting is provided "AS IS" with no warranties or guarantees, and confers no rights.

    Sunday, January 14, 2018 2:22 PM
  • Exact the same behaviour in my case running Exchage within the Hyper-V VM with Windows Server 2012 r2.

    I managed to deinstall the Update, and everything worked again as before. Would be nice to release a new version of the update KB4057903 without this bug.

    Thanks

    Sunday, January 14, 2018 2:39 PM
  • Would be nice to release a new version of the update KB4057903 without this bug.

    Thanks

    You can provide this feedback over here on uservoice.

    https://windowsserver.uservoice.com/forums/304618-installation-and-patching?category_id=141228

     

     

     



    Regards, Dave Patrick ....
    Microsoft Certified Professional
    Microsoft MVP [Windows Server] Datacenter Management

    Disclaimer: This posting is provided "AS IS" with no warranties or guarantees, and confers no rights.

    Sunday, January 14, 2018 2:44 PM
  • Done. Thanks for the suggestion.
    Sunday, January 14, 2018 4:10 PM
  • You're welcome.

     

     



    Regards, Dave Patrick ....
    Microsoft Certified Professional
    Microsoft MVP [Windows Server] Datacenter Management

    Disclaimer: This posting is provided "AS IS" with no warranties or guarantees, and confers no rights.

    Sunday, January 14, 2018 4:15 PM
  • I have voted for your post
    Sunday, January 14, 2018 6:11 PM
  • I can also confirm this bug. Running exchange 2013r2 CU19 on HyperV 2012R2. After installing KB4057903 server becomes unresponsive. I had to restore from backup.
    Sunday, January 14, 2018 9:54 PM
  • This is definitely a bug. Running exchange 2013 in Hyper-V and the server hangs when rebooted.
    Monday, January 15, 2018 2:26 AM
  • This is definitely a bug. Running exchange 2013 in Hyper-V and the server hangs when rebooted.

    You can call it in here

    https://support.microsoft.com/en-us/gp/contactus81?Audience=Commercial&wa=wsignin1.0

    but I'd also try installing from a clean boot

    https://support.microsoft.com/en-us/help/929135/how-to-perform-a-clean-boot-in-windows

    or you can also provide this feedback over here on uservoice.

    https://windowsserver.uservoice.com/forums/304618-installation-and-patching?category_id=141228




    Regards, Dave Patrick ....
    Microsoft Certified Professional
    Microsoft MVP [Windows Server] Datacenter Management

    Disclaimer: This posting is provided "AS IS" with no warranties or guarantees, and confers no rights.


    Monday, January 15, 2018 2:27 AM
  • Same Issue here. Also only Exchange server affected.

    The update should be removed by Microsoft and reapplied after a fix. How can we monitor when this will be done => replacement is available?


    Monday, January 15, 2018 9:52 AM
  • KB4057903 updates a dll (vmstorfl.sys) from the integration services.

    Simply solution: 

    - deactivate Exchange services
    - install KB4057903
    - reboot
    - reactivate Exchange services

    Same behavior as on updating integration services on an exchange server. Worked perfect for me on several Exchange VMs. 


    • Edited by rennstrecke Monday, January 15, 2018 4:45 PM
    • Proposed as answer by rennstrecke Monday, January 15, 2018 4:48 PM
    Monday, January 15, 2018 4:44 PM
  • How did you deactivate and reactivate Exchange services ?

    Thank you.

    Monday, January 15, 2018 7:28 PM
  • Hi eeryelf,

    We have reported this topic to related team, if there's any feedback there, we'll post on forum as soon as possible.

    Best Regards,

    Anne


    Please remember to mark the replies as answers if they help.
    If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com.

    Tuesday, January 16, 2018 8:57 AM
    Moderator
  • Just to add to this...

    KB4057903 fails to install on many of the 2012 R2 servers that we have, not just Exchange. As others have mentioned, we just get the dreaded "failed to install updates" message upon reboot and there is little info about.

    I have noticed the same behaviour in the past with other integration service update releases for 2012 R2

    So far, we have only managed to install this on one 2012 R2 VM successfully, the rest have all failed.

    Thanks

    Matt

    Wednesday, January 17, 2018 11:12 AM
  • This has been an issue with Integration Services updates and Exchange for some time.

    See this post from back in 2016

    (also includes steps to resolve it)

    http://blog.scng.si/exchange-server-vm-becomes-unresponsive-while-updating-hyper-v-integration-services/


    Andrew

    Thursday, January 18, 2018 12:56 PM
  • After installing the update, my 2012 R2 with Exchange 2013 on Hyper-V appeared to hang at "Getting Windows Ready" with the circling animation stopped.

    I left it intact for approx. 40 minutes... and the Windows startup just resumed itself. The update was installed successfully. All Exchange services are running.

    Thursday, January 18, 2018 4:57 PM
  • I try this solution and it's did not work for me :

    You can make this a bit easier by just changing the “Microsoft Exchange Active Directory Topology Service” to manual (or disabled). Most Exchange services depend on this, so they won’t start without it.

    I left it intact for approx. 40 minutes... and the Windows startup just resumed itself but the server was not respond to ping at all and evrthing was very very slow ! So i restore from backup bedore the update to make it work again...

    Some one have a real working solution ?

    Thank you.

    Thursday, January 18, 2018 5:38 PM
  • I try this solution and it's did not work for me :

    Some one have a real working solution ?

    Thank you.

    Yes...
    Disable ALL the exchange service
    Reboot
    Apply Update
    Reboot
    Enable Exchange Services
    Reboot

    Or if you have already applied the update and your server is not responsive
    Boot from Windows ISO
    bcdedit to enable boot menu
    boot into safe mode
    disable all exchange services
    reboot
    apply updates
    reboot
    enable exchange services
    reboot

    I have used this successfully on Exchange 2013 and 2016 on both Hyper-V 2016 and Hyper-V 2012R2

    I have a script to enable and disable all the exchange services now, so it only takes a second.

    Andy


    Andrew

    • Proposed as answer by Champac Friday, January 19, 2018 5:30 PM
    Thursday, January 18, 2018 11:23 PM
  • Andrew, Can you send me this script by email or where i Can find IT?

    Thank a lot.

    Friday, January 19, 2018 2:56 AM
  • Hi,

    Ex2013:
    http://www.oakeley.com.au/Ex2013-Disable.txt
    http://www.oakeley.com.au/Ex2013-Enable.txt

    Ex2016:
    http://www.oakeley.com.au/Ex2016-Disable.txt
    http://www.oakeley.com.au/Ex2016-Enable.txt


    Andrew

    Friday, January 19, 2018 3:10 AM
  • Andrew,

    I just apply your solution on my Hyper-V Windows Server 2012R2 running Exchange 2012 CU19 and evrything work like a charm!

    Thank a lot!

    Microsoft please stop to do update like that !!

    Friday, January 19, 2018 5:31 PM
  • Andrew you are my hero! ;) You solution works perfect!

    • Edited by medusa666 Saturday, January 20, 2018 1:32 PM
    Saturday, January 20, 2018 1:28 PM
  • Andy, thanks a lot for the solution and for the two scripts!

    BTW, I had to add ".exe" to the sc command in order to run it within PowerShell.

    Regards, José

    Sunday, January 21, 2018 2:30 PM
  • Having the same issue as everyone else running a Hyper-V 2012 Exchange Server here. Exchange 2013 CU 18 to be exact.

    Andrew's recommendation of disabling the exchange services & applying the update works perfectly, but for those who are already stuck in this, booting off an ISO can be a bit of a pain.

    In my experience, i was able to open a CMD window via task manager immediately after logging in & from there you can run the commands to display the boot manager on the next reboot.

    Command to enable : bcdedit /set {bootmgr} displaybootmenu yes

    Command to disable : bcdedit /set {bootmgr} displaybootmenu no

    This will allow you to boot into safe mode easier & quicker. 

    PLEASE MICROSOFT FIX THIS ISSUE! I spent way to long this morning fighting with this.


    Sunday, January 21, 2018 6:28 PM
  • Having the same issue as everyone else running a Hyper-V 2012 Exchange Server here. Exchange 2013 CU 18 to be exact.

    Andrew's recommendation of disabling the exchange services & applying the update works

    I would love to take credit for this, but the solution was already published by someone else back in 2016.

    http://blog.scng.si/exchange-server-vm-becomes-unresponsive-while-updating-hyper-v-integration-services/

    All I did was put together some scripts to save a lot of time disabling and re-enabling services

    http://www.oakeley.com.au/exchangescripts.htm

    I am glad this has helped people though


    Andrew

    Monday, January 22, 2018 9:51 AM
  • Hi!

    We have faced an other issue with this update.
    We have several W2K12R2 Domain Controllers running as Hyper-V VM.
    These DCs have more than one NIC. So we have configured the DNS Server "Interfaces" Setting to
    "Listen on only the following IP addresses" an selected the right one only.

    After applying the Update KB4057903 the DNS Server "Interfaces" Setting was reset do default
    "Listen on all ip addresses".

    We were able to reproduce that in a lab environment where we have restored an affected DC
    from backup before applying KB4057903, checking the DNS Server Settings and applying KB4057903.
    The Setting was changed from "Only the following IP addresses" to "All IP addresses".

    Does someone else had this kind of behavior?

    br, Rainer

    Monday, January 22, 2018 12:53 PM
  • Just to add to this...

    KB4057903 fails to install on many of the 2012 R2 servers that we have, not just Exchange. As others have mentioned, we just get the dreaded "failed to install updates" message upon reboot and there is little info about.

    I have noticed the same behaviour in the past with other integration service update releases for 2012 R2

    So far, we have only managed to install this on one 2012 R2 VM successfully, the rest have all failed.

    Thanks

    Matt

    I have the same problem with this KB4057903 on 1 Win2012 R2 VM. No Exchange services are installed on this machine. The weirdest issue is some other VMs, also W2012r2 and running on the same cluster host can install this KB normally. These are literally no difference in the VMs which are running the same type of workloads and applications. All integration services are offered for both VMs by the host.

    Voted in uservoice.

    Wednesday, January 24, 2018 2:34 AM
  • Server 2012 R2/Exch 2016 RTM on Server 2012 R2 host

    Simiarly experience - installed update, after 10 - 15 min, server came up minus network adapter. Rolled back to checkpoint taken before update. Same thing second time, but while researching (and finding this thread) the server rebooted itself after being up about 40 min, and everything was running - Exchange services, cluster role, all auto start services.

    Installing the rest of the updates...then if all goes well, CU 8.

    Wednesday, January 24, 2018 7:19 PM
  • Why hasn't this bad patch been pulled? It's CLEARLY causing a lot of issues out there. It bit us at 3 sites now before I found the issue and pulled approval on it. Microsoft: Get your house in order. Test your patches better and respond better when problems are found - this patch should have been IMMEDIATELY pulled due to these issues until a new patch can be created that doesn't cause serious issues like this.The fact that this story repeats itself over and over and over almost every month is proof positive that Microsoft has lost sight of what is important. Quality and customer service are no longer priorities, only money. For shame!

    Friday, January 26, 2018 12:49 PM
  • This issues has been around for quite a while, when installing new integration services on a VM running Exchange the services collide during bootup and the NIC driver fails to load.  It always amounts to a very long wait for 'we couldn't complete the updates undoing changes do not turn off your computer'.  

    There must be some way for Microsoft to detect that the virtual server is running Exchange and delay the start of Exchange Services?  It must be happening everywhere as I have seen it 1/2 dozen times at my clients.  Always a nightmare.

    Please address this oversight Microsoft Hyperv Team

    Friday, January 26, 2018 8:32 PM
  • This issues has been around for quite a while, when installing new integration services on a VM running Exchange the services collide during bootup and the NIC driver fails to load.  It always amounts to a very long wait for 'we couldn't complete the updates undoing changes do not turn off your computer'.  

    There must be some way for Microsoft to detect that the virtual server is running Exchange and delay the start of Exchange Services?  It must be happening everywhere as I have seen it 1/2 dozen times at my clients.  Always a nightmare.

    Please address this oversight Microsoft Hyperv Team

    Better to post feedback over here on uservoice.

    https://windowsserver.uservoice.com/forums/304618-installation-and-patching

     

     



    Regards, Dave Patrick ....
    Microsoft Certified Professional
    Microsoft MVP [Windows Server] Datacenter Management

    Disclaimer: This posting is provided "AS IS" with no warranties or guarantees, and confers no rights.

    Friday, January 26, 2018 8:41 PM
  • Why is the non-Microsoft site UserVoice better than an official Microsoft forum? Has Microsoft senior management indicated that they actually listen to this third party forum and not their own?

    Edit: Oh, I see. UserVoice is a PAID subscription service.  I would have to PAY a third party company to add my comment that Microsoft has a broken patch...... And they apparently - do something with my comment. And just what is your financial stake in UserVoice? I can't imagine anyone advocating for it unless they had some sort of financial incentive.

    • Edited by Walt Reed Friday, January 26, 2018 8:59 PM
    Friday, January 26, 2018 8:52 PM
  • Yes, correct. Postings here in forums are not categorized or arranged, other than by general technology, and will just eventually roll off in to nothingness

    https://blogs.technet.microsoft.com/askcore/2015/05/20/invitation-to-provide-feedback-through-uservoice/

     

     



    Regards, Dave Patrick ....
    Microsoft Certified Professional
    Microsoft MVP [Windows Server] Datacenter Management

    Disclaimer: This posting is provided "AS IS" with no warranties or guarantees, and confers no rights.


    Friday, January 26, 2018 8:54 PM
  • From the page you linked to:

    Remember, these sites are for feature suggestions and ideas only.

    Bug reports are not feature suggestions or general ideas. In other words, UserVoice is the wrong forum for reporting bugs like - THIS PATCH BREAKS STUFF.

    Friday, January 26, 2018 9:17 PM
  • From the page you linked to:

    Remember, these sites are for feature suggestions and ideas only.

    Bug reports are not feature suggestions or general ideas. In other words, UserVoice is the wrong forum for reporting bugs like - THIS PATCH BREAKS STUFF.

    Then you'll do better to call it in here.

    https://support.microsoft.com/en-us/gp/contactus81?Audience=Commercial&wa=wsignin1.0

     



    Regards, Dave Patrick ....
    Microsoft Certified Professional
    Microsoft MVP [Windows Server] Datacenter Management

    Disclaimer: This posting is provided "AS IS" with no warranties or guarantees, and confers no rights.

    Friday, January 26, 2018 9:20 PM
  • Thanks for the scripts - I had to change a couple services because their startup state was different on my server

    FYI there's a type in the disable.txt

    sc.exe config MsceomplianceAudit start=disabled
    

    Thursday, February 1, 2018 9:30 PM
  • I disabled the startup of the Exchange services (with Andrew's script) before I updated the second Exchange server. I also rebooted so the services would not be running while the update installed before the reboot, in case it mattered. The second server updated without the 40 min. delay.
    Thursday, February 1, 2018 9:33 PM
  • FYI there's a type in the disable.txt


    Fixed. Ta

    Andrew

    Thursday, February 1, 2018 11:02 PM
  • We had another interesting problem with this update. We have 2 DC's - one on premise and a secondary as  VM in Azure. Our on premises VM's are replicated to Azure Site recovery. After applying the update the Replication for our DC failed. Only the DC the other VM's replicated fine. We checked with VVSADMIN LIST WRITERS and the NTDS failed. We removed the update which fixed the replication and Writers error but what came next was even better. Our AD had now journal warp errors. So we fixed the journal warp errors and after we had not SYSVOL and NETLOGON shared folders on both DC's. We performed a restore of the sysvol from backup performed FRS D4 on on premise DC  Syvol and netlogon got shared again on the VM in Azure performed FRS D2.

    What a bug. And still not fixed. 

    Sunday, February 18, 2018 12:53 PM
  • Just to add to this...

    KB4057903 fails to install on many of the 2012 R2 servers that we have, not just Exchange. As others have mentioned, we just get the dreaded "failed to install updates" message upon reboot and there is little info about.

    I have noticed the same behaviour in the past with other integration service update releases for 2012 R2

    So far, we have only managed to install this on one 2012 R2 VM successfully, the rest have all failed.

    Thanks

    Matt

    I have the same problem with this KB4057903 on 1 Win2012 R2 VM. No Exchange services are installed on this machine. The weirdest issue is some other VMs, also W2012r2 and running on the same cluster host can install this KB normally. These are literally no difference in the VMs which are running the same type of workloads and applications. All integration services are offered for both VMs by the host.

    Voted in uservoice.

    The latest KB4072650 also for HyperV Integration Components broke my other server which is an almost exact replica of the one that had problems with KB4057903. Symptoms after install was the same - no network, sluggish to almost no response.

    The curious thing is other VMs in the same HyperV cluster are able to take these 2 KBs normally.
    Both the problematic VMs do not have Exchange services running.

    Friday, February 23, 2018 6:44 AM
  • Same problem again! :-(

    In my case, it's a VM with Exchange on Windows Server 2012 R2, see above.

    I hope I don't have to disable all Exchange services, install the update, and enable the services again for all future Windows Updates. I will wait with this procedure this time since KB4072650 is optional.

    Saturday, February 24, 2018 4:10 PM
  • Just had the same issue on 2012 R2 Exchange 2013 CU13 VM after installing KB4072650, on reboot the server just hung on getting windows ready. Luckily I'd taken a checkpoint so was able to roll back.



    • Edited by Bristow2013 Sunday, February 25, 2018 10:33 AM
    Sunday, February 25, 2018 10:33 AM
  • I work for an MSP and we have seen this issue A LOT!! Not just on Exchange, but other servers too. RD Session Hosts, Domain Controllers. You name it. All Server 2012 R2.

    Not only was KB4057903 affected but now we seem to have the EXACT same issue with KB4072650 as well.

    Rather than the stock, corporate line of "Go to UserVoice", why not at least make some sort of vain attempt at trying to offer the community some advice here? It would seem as though nothing has been done between KB4057903 and KB4072650 being released.

    The silence in here is absolutely deafening Microsoft...

    Tuesday, March 13, 2018 3:05 PM
  • Same problem with me, the update KB4072650 try 3 time to install is self, after the tird , my server as been rolling back.

    Question: Do I need to install the KB4057903 before installing KB4072650, because I hide this marvellous update KB4057903. (Sarcasm)

    I realy do not like patching tuesday!!!!

    Wednesday, March 14, 2018 8:04 PM
  • Hi!

    Configuring all Exchange services as disabled, let me install the updates without issues.


    Cristian L Ruiz

    Friday, March 16, 2018 2:35 PM
  • I work for an MSP and we have seen this issue A LOT!! Not just on Exchange, but other servers too. RD Session Hosts, Domain Controllers. You name it. All Server 2012 R2.

    Not only was KB4057903 affected but now we seem to have the EXACT same issue with KB4072650 as well.

    Rather than the stock, corporate line of "Go to UserVoice", why not at least make some sort of vain attempt at trying to offer the community some advice here? It would seem as though nothing has been done between KB4057903 and KB4072650 being released.

    The silence in here is absolutely deafening Microsoft...

    Yes, I have some VM affected by KB4077903, then another VM by KB4072560. Same symptoms and both KB are related to Hyper V integration components. None of my VMs have exchange services installed.

    Really can't get my head around why the KB can be installed on some VMs, and not others.
    They are really similar running all the same services.

    So tired of patch tuesdays recently!

    Thursday, March 22, 2018 7:43 AM
  • I too am growing weary of spending whole days fixing things every month because Microsoft couldn't be bothered to test updates before they push them out. (In a small business, it's cost prohibitive to have an entire test lab of the production network to test everything on first). At the very least, a patch to a Server product should be tested on machines with other supported Microsoft products installed and not released if it breaks any of them. I got lucky this time, after several failed reboots trying to fix this one, I finally resigned myself to having to restore a backup which I did in parallel to a different location. At some point during the hours while that was restoring, the thing finally sorted itself out and came back up with the patch failed. I was happy to leave it not installed and unapproved on my WSUS.

    Somewhat related to this, I dumped my Lumia 950XL and started using an iPhone again, after years on Windows, because Microsoft updated my phone to the point where it is unstable and won't run for more than a couple of days without requiring a reboot, which it will do sometimes in the middle of a phone call. It's getting to the point that when I hear about a Microsoft product that is working well, my first thought is "give them time, it just hasn't been updated enough yet."

    My entire career has been built around network administration involving Microsoft products. For the first time since my days in college, I have begun to feel insecure about that choice. I remember the days of "nobody ever got fired for buying IBM" and we know where that went. It seems Microsoft is headed down that same path.






    Thursday, March 22, 2018 2:41 PM
  • At least with Exchange servers, it seems that there are two reboots in the patch process, and the first reboot is without virtual network adapters installed, so the Exchange services delay the boot process because there is some kind of wait period for finding the network adapters. Thus, if the services are disabled, there is no delay. Could that be the common thread?
    Friday, March 23, 2018 1:37 PM
  • Hi Dave,

    I expiereince the same problem with KB4072650 and KB4057903.

    A clean boot has not helped on a server platform and is just a waste of time.

    Regards

    Mick

    Monday, March 26, 2018 7:42 AM
  • Thanks and cudos to Andrew!

    we had a problem with one of our two Exchange servers not booting up correctly and while Bing-ing for an answer i came upon this thread and possible solution to our problem. while we had one server failing we also suspected the other one would fail as well... this thread looked very promising to us and i can now say it is :-)

    Went through the process according to Andrews suggestion (running the provided scripts for disable/enable) on the failing server and managed to save hours of restore and loss of data on that one.

    as i knew the solution worked on the failing one i started the process prior to rebooting the server i suspected would fail as well:
    1. copied all scripts i needed to c:\temp on the VM (also just in case, make sure you have access to local administrators account - i used it for all the following steps)
    2. ran "bcdedit /set {bootmgr} displaybootmenu yes" in elevated promt prior to reboot of VM. the reboot took forever as it tried to install the updates
    3. during reboot of VM, i selected Safe Mode (you need to press F8 in order to enter advanced boot) and noted that updates were reverted during start, this then initiated a new reboot of VM and here you need to enter Safe Mode once again in order to move on - or the VM will yet again try to install the updates
    4. disabled all Exchange services (see Andrews script), then rebooted the VM again
    5. installed the updates from Microsoft, rebooted VM, double checked for more updates (i would suggest installing them Before moving on, they might require another rebbot of VM)
    6. enabled all Exchange services (yet again see Andrews script), then rebooted the VM again
    7. verified all functions to be back and Active, no more updates available at this time so i reported to everyone
    8. finally reverted my change to bootmenu "bcdedit /set {bootmgr} displaybootmenu no" from elevated prompt

    we are finally up to date on Windows Updates and have an Exchange environment up and running.

    yet again, thanks are in Place for the aid provided by this thread and Andrew in particluar for the scripts :-)

    • Proposed as answer by RonnyDanger Friday, October 5, 2018 9:29 AM
    Wednesday, March 28, 2018 9:55 AM