locked
Upgrading Tanjay to R2 : Freeze in authentication step RRS feed

  • Question

  • Hy,

    I have implemented in my existing OCS R1 standard architecture an OCS R2 standard server (On hyper-v). And I would like to update my CX700. So I downloaded the UCUpdates and upload the CAB file into my Device Updater. Then I approve the firmware for the CX700. Finally I restart the phone by unplugging-plugging the power.

    If i sign with a user which is in on the OCS R2, it start and freeze during the authentication step and it is repeating. I have try to purge the images, but no effect.

    EDIT :
    My last update status is 0x2ee7/0
    This blog says this error mean "12007 ERROR_INTERNET_NAME_NOT_RESOLVED (0x2ee7) The server name could not be resolved. "

    On this post Jens suggest to "Use wbemtest and connect to WMI and check the MSFT_SIPUpdatesServerSetting instance and the properties InternalUpdatesDownloadURL and ExternalUpdatesDownloadURL"
    After running the wbemtest, it showed that the InternalUpdatesDownloadURL and the ExternalUpdatesDownloadURL were both null...

    The IIS log indicate only access to the Address Book : GET /Abs/Int/Handler/D-0b99-0b9a.lsabs There is no indication of access to the updates

    So what I have done wrong ? How can I diagnostic my issues (Where can I find the log of the Tanjay ?) ?

    Thanks
    Wednesday, February 18, 2009 11:45 AM

Answers

  • Hi,

    As I had many issue with my current OCS implementation to update my Tanjay. I have install a fresh (virtual) standard OCS R2, with a new AD, new DNS...
    So on my dedicated implementation of OCS, i have an AD, DNS, WINS, NTP, DHCP, CA (How to make the Root CA certificate available for Office Communicator 2007 Phone Edition?).
    Then I add the update : 1.0.522.101 (As my Tanjay was 1.0.522.34) and 3.5.6907.0

    After a hard reset my Tanjay have been updated !

    Then I plugged the updated Tanjay into my current OCS implementation and it works.

    Thanks for you help Drago !
    • Marked as answer by BougRun Tuesday, March 10, 2009 7:13 AM
    Tuesday, March 10, 2009 6:58 AM

All replies

  • Do you have the proper NTP and other DNS records setup?
    Thursday, February 19, 2009 4:21 AM
  • Thanks for your answer,

    I have check form a computer inside the domain :
    The NTP record with a net time, it works.
    The DNS record with a nslookup, it works too.

    What other tests can I run ?
    Thursday, February 19, 2009 6:42 AM
  •  You can view Tanjay's log after uploading it to the OCS server. Apparently you cannot do that at the moment.

    1. What is the current version installed on the device?
    2. Currently, can you sign-in as a user homed on R1 front end?
    3. Missing UpdatesDownloadURL is bad. We can populate with what suppose to be but probably something went wrong during the installation...

    Drago
    Thursday, February 19, 2009 12:37 PM
  • 1. The version of my CX700 is 1.0.522.34 (1.23)

    2. Yes, i can sign-in as an R1 user. But I have to reset it first, to be able to sign-in as a different user

    3. What is the url do I have to enter in the InternalUpdatesDownloadURL ? I knows the update files are store in this folder : C:\Program Files\Microsoft Office Communications Server 2007 R2\Web Components\DeviceUpdateFiles\DeviceUpdates\UCPhone\Polycom\CX700\A\ENU\3.5.6907.0\CPE.
    Is this url correct : https://R2-Front-End-Server/AutoUpdate/Int/Handler ?

    Beside the Device Updater Console, is there some settings elsewhere ?

    Thanks

    Friday, February 20, 2009 7:23 AM
  • 1. You cannot update anything below 1.0.522.101 (1.23) via R2 Update services

    ***Correction*** There is a interm update (1.0.522.103 (1.23) along with 3.5.6907.0 (1.23) which should allow earlier version to update to R2, but I have not seen it in action... In your case -the missing URL's are the most probable cause.

    2. 1.0.522.101 (1.23) is the minimum required version in order to sign-in to R2 environment.

    3. the correct URL's are as follow:

        a. for InternalUpdatesDownloadURL: https://FQDN-front-end/RequestHandler/ucdevice.upx 
        b. for InternalUpdatesStoreURL: http://FQDN-front-end/DeviceUpdateFiles_Int

    What you need to do is:

    1. Fix the missing URL issue (you still might have problems because those should be populated by default during the installation)
    2. Upload 1.0.522.101 (1.23) update to the R2 server (do not approve it!)
    3. Enter your device as "test" device
    4. Create DNS record for ucupdates-r2 with IP address your R2 front end server
    5. Perform hard reset on your device
    6. Leave it alone - (it should take 1.0.522.101 (1.23) update within 15 min. and reboot)
    7. Once updated to 1.0.522.101 (1.23), remove it from the "test devices" tab.
    8. Either sign-in with R2 account and wait (the device must be idle for 10 min. to begin the update installation) or do another hard reset and wait.

    Drago

    Friday, February 20, 2009 2:11 PM
  •  >>The DNS record with a nslookup, it works too

    Which DNS record did you resolve correctly?  I ran into a similiar situation.  You need an A record called ucupdates.<domain>.<whatever>

    Unfortunately, I also needed a WINS server or I had to be on the same Subnet as my internal CA server sat.
    Friday, February 20, 2009 6:26 PM
  • UcUpdates.sipdomain.com is for R1 Update services. R2 uses UcUpdates-r2.sipdomain.com.

    Having both in mixed environment could lead to confusion among devices...

    Drago
    Friday, February 20, 2009 7:56 PM
  • I have another installation (for the test) of OCS Standard Edition R2, and I check the InternalUpdates URL, it is also null. It is strange that I have the same issue on two completely different installation. Here is a screenshoot of the original MSFT_SIPUpdatesServerSetting instance, perhaps I misinterpreted it :
     Free Image Hosting at www.ImageShack.us

    But I have fix the URL issue on my production Front-End R2 like you suggest. I have upload the update and add my device as test. I had already created the ucupdates-R2 record. When I reset the phone and wait for an updates. Nothing happens, and I have the same error : "Last update status : 0x2ee7/0". I have try to logon as an R2 user, like before it freeze during the authentication step and it is repeating. And my IIS logs does not contains anything about access to update site.


    To respond to griffter when I do a nslookup of ucupdates-R2.mydomain.fr, I have this :

    Serveur :  myDNSserver.uc.mydomain.fr
    Address:  192.168.100.4

    DNS request timed out.
    timeout was 2 seconds.
    DNS request timed out.
    timeout was 2 seconds.
    Réponse ne faisant pas autorité :

    (non-authoritative answer)

    DNS request timed out.
    timeout was 2 seconds.
    Nom :    ucupdates-R2.mydomain.fr
    Address:  192.168.100.17


    I think this is because I have two domain : lan.mydomain.fr and uc.mydomain.fr
    The uc.mydomain.fr is a DMZ with all the OCS and exchange server. (All my updating test has been done by connecting the phone directly inside the DMZ.)
    And the DNS for mydomain.fr is not local, it is on a hosting services. So all the records.mydomain.fr has to be created on the hoster DNS.

    So will it help if I create locally on my DNS server a mydomain.fr zone ?

    Perhaps as I approved the 3.5.6907.0 updates, it cannot dowload the 1.0.522.101 updates. How can I disapprove/reject the 3.5 updates ?

    Thanks for your help
    Monday, February 23, 2009 7:20 AM
  • Most (or all) of your troubles are coming from misconfigured DNS. This includes 0x2ee7/0 (ERROR_INTERNET_NAME_NOT_RESOLVED)

    Get with me via this email address: drago at alltel.net


    Drago
    Monday, February 23, 2009 1:21 PM
  • Hi,

    As I had many issue with my current OCS implementation to update my Tanjay. I have install a fresh (virtual) standard OCS R2, with a new AD, new DNS...
    So on my dedicated implementation of OCS, i have an AD, DNS, WINS, NTP, DHCP, CA (How to make the Root CA certificate available for Office Communicator 2007 Phone Edition?).
    Then I add the update : 1.0.522.101 (As my Tanjay was 1.0.522.34) and 3.5.6907.0

    After a hard reset my Tanjay have been updated !

    Then I plugged the updated Tanjay into my current OCS implementation and it works.

    Thanks for you help Drago !
    • Marked as answer by BougRun Tuesday, March 10, 2009 7:13 AM
    Tuesday, March 10, 2009 6:58 AM
  • Hi all,

    I´m also having some troubles updating our tanjay phone (from 1.0.452.0) to a R2 firmware. The phone reports 0x0/200 in the last update status, it also receives the ucdevice.upx file (traced with wireshark) but it never downloads the cpe.nbt file.

    I have also run wbemtest on both frontend servers and InternalUpdatesStoreURL & InternalUpdatesDownloadURL are both empty.  But when I run  lcscmd /web /action:ListWMISettings /poolname:<our poolname> and open the html file which was generated, the InternalUpdatesStoreURL & InternalUpdatesDownloadURL are both present.

    We don´t have a R1 server anymore and tanjays with firmware 1.0.552.101 updated without a problem to the R2 firmware (3.5.6907.0). When performing an nslookup of the update URL I get the correct response.

    Did somebody ever manage to update any firmware lower then 1.0.552.0 with only a R2 environment to the R2 firmware?

    Any help will be appreciated.
    Best regards
    Tuesday, April 7, 2009 10:54 AM
  • I'm expecting 55 units this Friday or early next week. I will post my experince once have chance to play with them.

    Drago
    Tuesday, April 7, 2009 9:42 PM
  • Hi Dragot,

    it would be nice to know what firmware version you have on the new phones. Did you had any problems upgrading the firmware?

    Best regards
    Wednesday, April 15, 2009 5:45 AM
  • The shipment was delayed and should arrive Thursday this week. I will let you know...

    Drago
    Wednesday, April 15, 2009 11:25 AM
  • The shipment finally arrived today. Need les to say, I was eager to connect at least one because of my previous problems. For my big relief, the firmware was 1.0.522.71

    One thing I've notices immediately - even thought the units are brand new, it went straight to login screen. This is - the screen calibration was not present. This pick up a red flag immediately. When checked the settings, the Update Status was 0x2ee2/0

    This, along with the missing calibration screen, made me do "hard reset" immediately. After boot, the calibration screen was present. So, I left the unit alone, went out for for a smoke and when got back, the phone was 3.5.6907.0 (1.23) i.e. updated.

    I currently have OCS R2 EE in Production environment. Best Practices analyzer runs without errors. UcUpdates-r2.domain.com DNS record is present.

    By default, the Client Version Filter is set to update everything above 1.0.522.* which indeed is the case here.

    *** I attempted to login with a second device that was not updated to R2 environment just to see what happens and it froze on Auth. step. I would say that the workaround for devices below 1.0.522.* is - one should have OCS 2007 ST and UC Update services on virtual environment for upgrading to version above 1.0.522.* or... insist that the supplier ships units with R2 acceptable firmware. Frankly, I highly doubt Microsoft will do anything about it. We have seen this before.

    Drago

     

    Friday, April 17, 2009 10:19 PM
  • Hi Dragot,

    thanks for your response. I´m in contact with our supplier which has a R1 running, although I don´t know if he has the update server deployed.

    Best regards
    Monday, April 20, 2009 11:21 AM
  • Monday, April 20, 2009 2:09 PM
  • I will be damned! Since the upgrade was running so smooth, I decided to delete the "ucupdate" DNS record and keep only "ucupdates-r2" since we're moving all users to R2. All of a sudden, Tanjay's have stopped updating. I thought I'm losing my mind...

    After all day sniffing and tracing - looks like if the firmware is 1.0.522.101 (1.23), the DNS query for ucupdates-r2 is build in. With other words, Tanjay with FW 1.0.522.71 (1.23) was looking for... ucupdates.domain.com

    I created ucupdates.domain.com pointing to my EE Pool and we're back in the game again.

    Drago

    Tuesday, April 21, 2009 8:39 PM