none
Random "The Microsoft Exchange Administrator has made a change that requires you quit and restart Outlook" message RRS feed

  • Question

  • I've looked and looked again and cannot find any solution to this issue...There are plenty of posts related to this error message however none of the resolutions work for me.

    We have a cluster of 3 MS Exchange Server 2010 SP2 RU2 machines with CAS, HT and Mailbox roles installed on all 3. Those 3 machines form a DAG and are behind a CAS Array. This is done using a pair of  Riverbed Stingray Traffic Manager appliances.

    The above error message keeps popping up randomly on my users' workstations running Outlook 2010 with cache mode enabled (Outllok 2007 users do not have this issue). It pops up when nothing is being done to the platform...i.e. I'm not moving any mailboxes, not activating nor suspending any mailbox databases, not updating the rpcclientaccessserver attribute on any databases (that property is configured correctly on all my databases...). We have SP2 RU2 installed which means that the message should only pop up if the conditions in THIS ARTICLE are observed and none of them are when our users see the message...

    Next thing on my elimination process is the load balancers...I've read about some admins having this issue when using Barracuda LBs for their CAS array (see this thread). I thought about session persistence settings not working properly but after some testing I figured out it doesn't make any difference if I use 'Source IP' or 'Cookie Based' as my persistence method for various services (OWA, MAPI, etc). The end users still see the message popping up randomly. Can anyone confirm Source IP is better? Or shall I stick with Cookie based for OWA services? Also I'm offloading SSL on the load balancers but I fail to see how this would be a problem. It's configured OK as per MS' SSL offloading guidance.

    Those users use Public Folders and my PF store reside on one of our servers only (as we are in a DAG). Because we have all 3 roles installed on all 3 servers, users can potentially connect to either 3 (if we assume that session persistence is not working properly). Could this be an issue?

    The pop ups are so random that I'm running out of ideas on how to troubleshoot this...

    Has anyone got any idea at all?

    Other things I've tried or checked:

    - tested Outlook connectivity OK

    - modified pool and service settings on my load balancers to check if those triggered the pop up, to no avail (only when I switch the main virtual IP from one traffic manager to the other the message pops up, but that is to be expected!)

    - tried disabling the 'show Exchange messages' option on the client side (which is obviously not ideal...but the pop up still occurs anyway)


    Regards, Raphael

    Tuesday, June 12, 2012 3:04 PM

All replies

  • If you put a hosts file on a client and bypass the LB altogether, do you see the issue?
    Tuesday, June 12, 2012 3:26 PM
  • Thanks for getting back to me Greg.

    I will try so but it may take a day or two before the message pops up. Thanks for the suggestion, I'll update here with test result.


    Regards, Raphael

    Wednesday, June 13, 2012 7:19 AM
  • Hi Greg,

    I do face the same problem - running two node E14 SP2 RU2 DAG after moving 2600 mailboxes last weekend. We bypass the F5 already - still the same issue using Outlook 2003/2007/2010 clients where randomly choosen ones do get the restart outlook message (at least we couldn't find a pattern). Also clients with mailboxes moved three weeks ago are affected. Both E14 multirole (C/H/M) have a replica for PF. There is one Exchange 2003 two cluster node left with around 350 mailboxes. Could it be caused by some sort of Outlook Add-In?

    many thanks

    Thursday, June 21, 2012 11:10 AM
  • Me again - I do not know the process how the exchange server is telling Outlook to restart (=to produce above message) so my question might sound silly.

    What will happen if MailboxA does get this kind of information (hidden item or what ever) and this mailboxA is not used = nobody is starting Outlook to work against this mailbox. UserB opening his mailboxB has delegated permissions f.i. to open the calendar of mailboxA. Somehow UserB/MailboxB does get the information that mailboxA should get "restartet" and the message is displayed even if UsersB mailboxB has not been moved nor touched.

    Thursday, June 21, 2012 11:19 AM
  • I have left a laptop on for a number of days with a test account logged in to Outlook 2010 (with modified hosts file to bypass the load balancers, e.g. always hit the same node). The message never popped up. This can certainly be a clue however unfortunately it doesn't 100% mean that the issue comes from the LBs as some users get the message randomly with sometimes more than 2 weeks apart...

    Another thing that we thought maybe a problem is we have noticed that this problem occurs more for people with laptops with both Ethernet and Wifi adapters plugged in at the same time. Those users un-dock their laptop to go to a meeting for example, and the wireless connection kicks in, then dock it back in to their docking station and the wireless connections stays on, as well as their Ethernet one. This sometimes changes the state of their Outlook connection status from TCP/IP to HTTPS and vice versa. This also depends on if the office LAN is running slow or not at the time. But surely I would have thought switching client IP address all the time may be an issue? Does anyone know if this could be a problem when connecting to a CAS Array/DAG Exchange platform?

    On the other hand I have tested having both connections on on that test laptop and the message never popped up. But yet again I was bypassing the load balancers.

    Their RPC over HTTPS Exchange Proxy Settings are all set with:

    - On fast networks connect using HTTP first then TCP/IP - disabled

    - On slow networks connect using HTTP first then TCP/IP - enabled

    I have a case open with our load balancer vendor and their Exchange expert is supposed to get in touch with me to review all the settings and Exchange config. I'll update this ticket if we can find more clues, or ideally, a solution to this issue!

    Hi Greg,

    I do face the same problem - running two node E14 SP2 RU2 DAG after moving 2600 mailboxes last weekend. We bypass the F5 already - still the same issue using Outlook 2003/2007/2010 clients where randomly choosen ones do get the restart outlook message (at least we couldn't find a pattern). Also clients with mailboxes moved three weeks ago are affected. Both E14 multirole (C/H/M) have a replica for PF. There is one Exchange 2003 two cluster node left with around 350 mailboxes. Could it be caused by some sort of Outlook Add-In?

    many thanks

    tiZ.A: Thanks for giving me some sanity back :-) I'm not the only one to whom this happens! do you know that moving mailboxes between databases is one of the legitimate reasons why this message would pop up for? (ref: http://blogs.technet.com/b/exchange/archive/2011/01/24/3411868.aspx)  The strange thing is that for you it doesn't seem to happen for ALL your mailbox users, but just random ones? Also we only have noticed this issue with Outlook 2010 clients ONLY. No 2003/2007 ones.


    Regards, Raphael




    • Edited by Sm0keN Thursday, June 21, 2012 12:04 PM
    Thursday, June 21, 2012 11:59 AM
  • Hi Raphael,

    I did face the same again at another customer while the notebook was switching networks and therefore protocols (RPC <=> HTTPs)!

    Friday, July 20, 2012 6:22 AM
  • I think I may have fixed out issue regarding this.

    I have read and re-read multiple articles on this "Exchange Administrator..." message over and over again and the one thing i'd missed every time was:

    "By preserving the mapping signature, we can minimize Outlook requiring a restart, as long as the following conditions are observed: * The public folder hierarchy server does not change. *"

    Basically, I thought that having one public folders database on one of our 3 mailbox servers was enough as when connected to the other two the client would get referred to the correct server...Well, now that we have one database per mailbox server and all replicas/folder structure and hierarchy match on all 3 servers, it seems the popup message issue has gone!

    Also, just to make sure, I have dedicated one role per server (i.e. removed the MBX and HT roles off our CAS boxes, and built dedicated MBX and HT servers).

    In other words, the load balancers weren't to blame IMO.

    It's been a week now and no one's had any pop up message issue. Fingers crossed it stays like that.

    Hope this helps you too tiZ.A


    Regards, Raphael

    Friday, July 20, 2012 9:54 AM
  • @Raphael: Till data has the problem problem come back again?
    Friday, November 9, 2012 12:01 PM
  • the same error happening to us as well.  This error is random and it just happened when the user switched the laptop from LAN to wireless. Collected the outlook logs and found really nothing.  Did anybody face this issue when the user switched from LAN to wireless? 

    How do we figure out the issue is caused by the network connections. Just a thing to remember we are using Basic authentication.

    I thinking whether changing it to NTLM would help? any idea?


    Wednesday, April 17, 2013 9:59 AM
  • Yes we have seen that switching to NTLM does resolve the issue. And yes it was like clock work for us to reproduce, just switch from wired to wireless and the issue will present itself.
    Friday, September 20, 2013 5:16 PM
  • The resolution in this case was to drill down to the properties of the Mailbox Database in ADSIEDIT & set the value of “msExchHomePublicMDB” to be blank. Afterwards, a restart of the Information Store Service resolved the issue.
    Monday, July 21, 2014 6:31 AM
  • I think I may have fixed out issue regarding this.

    I have read and re-read multiple articles on this "Exchange Administrator..." message over and over again and the one thing i'd missed every time was:

    "By preserving the mapping signature, we can minimize Outlook requiring a restart, as long as the following conditions are observed: * The public folder hierarchy server does not change. *"

    Basically, I thought that having one public folders database on one of our 3 mailbox servers was enough as when connected to the other two the client would get referred to the correct server...Well, now that we have one database per mailbox server and all replicas/folder structure and hierarchy match on all 3 servers, it seems the popup message issue has gone!

    Also, just to make sure, I have dedicated one role per server (i.e. removed the MBX and HT roles off our CAS boxes, and built dedicated MBX and HT servers).

    In other words, the load balancers weren't to blame IMO.

    It's been a week now and no one's had any pop up message issue. Fingers crossed it stays like that.

    Hope this helps you too tiZ.A


    Regards, Raphael

    This scenario is now documented here:

    https://blogs.technet.microsoft.com/vytas/2014/11/12/your-administrator-has-made-a-change-and-requires-you-to-restart-outlook/

    and more specifically, here:

    https://blogs.technet.microsoft.com/aljackie/2013/11/14/outlook-rpc-end-point-and-pf-the-microsoft-exchange-administrator-has-made-a-change-that-requires-you-quit-and-restart-outlook/


    Mike Crowley | MVP
    My Blog -- Baseline Technologies

    Monday, May 23, 2016 8:18 PM
  • We had the same error show up after a MS Exchange Schema update. The msExchHomePublicMDB attribute for our mailbox databases was referencing deleted/invalid Public Folders.

    After clearing the value and waiting for AD replication, the issue went away.

    You can view/clear this in ADSI edit. You'll need to connect to the Configuration context. The DN of the mailbox databases is something like this:
    CN=Databases,CN=Exchange Administrative Group (FYDIBOHF23SPDLT),CN=Administrative Groups,CN=domain,CN=Microsoft Exchange,CN=Services,CN=Configuration,DC=domain,DC=com



    Tuesday, April 9, 2019 3:22 AM