locked
"Your chatroom functionality may be limited due to an outage" client error after adding new Central site to SfB topology RRS feed

  • Question

  • Hello,

    recently I have updated our company Lync/Skype for Business topology with new Central site (2 SE FE paired, Mediation server, WAC, Persistent chat pool made of 2 Persistent chat nodes).

    So far according to my testing everything seems to be okay, but users from one particular site (Lync 2013 SE FE) reprorted, that they spotted error message within their SfB clients - "Your chatroom functionality may be limited due to an outage". It seems to be a false positive, since they can connect to chatrooms as they were used to before, its just that the error is bothering them simply out of being there.

    Currently there are 2 Persistent chat pools with the whole topology - one Lync 2013 (one node) and the one I recently added (SfB, 2 nodes). Site those "problematic" users are from has that Lync 2013 persistent chat configured for use (so they should not interact in any way with the new SfB one). All "production" chatrooms reside on Lync Pchat, I just created one for test purposes on that SfB pool.

    Now what I dont understand is, that when I take a look at client logs from a user that has this error message on his client, why there is:

    SIP/2.0 481 Call Leg/Transaction Does Not Exist

    ms-diagnostics: 1033;reason="Previous hop server component did not report diagnostic information";Domain="sipdomain.com";PeerServer="SFB_PCHAT_NODE_1";source="SFB_SEFE_NODE_1"

    (I replaced actual node/domain names with dummies)

    Why is users client trying to interact with that new SfB persistent chat pool when there is no direct (next hop) connection configured within the topology?

    Thank you for any insight,

    Tomas

    Wednesday, October 31, 2018 4:25 PM

All replies

  • Hi Tomas,

    Please check the next hop pool for the Lync 2013 FE pool, make sure the setting is true.


    According to the log you provided:
    SIP/2.0 481 Call Leg/Transaction Does Not Exist
    ms-diagnostics: 1033;reason="Previous hop server component did not report diagnostic information";Domain="sipdomain.com";PeerServer="SFB_PCHAT_NODE_1";source="SFB_SEFE_NODE_1", it seems an issue with SFB call forwarding, please make sure the PeerServer shows the server of Persistent Chat Server in your environment. 
     
    In addition, I suggest you could update the servers for Lync Server 2013 to the latest version, there are many issues has been fixed with the latest version.

    Best Regards,
    Evan Jiang


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


    Click here to learn more. Visit the dedicated forum to share, explore and talk to experts about Microsoft Teams.

    Thursday, November 1, 2018 2:20 AM
  • Hello,

    okay, please take a look at following topology (lot of blackouting but I think that basic overview is there:

    So, we have 5 Lync sites (deployed long time ago, up and running normally), and all are using Persistent chat pool (or, in this case, a server) that is deployed in the first (from top) site. 

    Then there is another pool, SfB one, that I added recently, which has its own Persistent chat pool that consist of 2 nodes).

    Now, as you can see, the "site where users with error are" has "PCHAT of first site on top" configured as default PCHAT pool. They are getting the error, but users homed on other sites (and their respective FE servers) dont see any errors.

    When I analyzed uccapi log from one user from that "site where users with error are", I saw the messages that I posted earlier - so, I saw that the client is for some reason referring to "PCHAT node that the error in logs refer to", which belongs to "new SfB site".

    And thats what I dont understand.

    (text in quotations refer to notes in the picture)

    Thank you for any advice,

    Tomas

    Thursday, November 1, 2018 12:37 PM
  • Hi Tomas,

    Are there any associated servers between the two sites, such as one Edge pool? 
    Did you do some changes in the site which users have the error? 

    Based on my experience, if you did not configure HA between two sites and do not have any associations, it should not have this issue. You could check the differences between the site which users have the error and the site which all works fine.

    Best Regards,
    Evan Jiang


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


    Click here to learn more. Visit the dedicated forum to share, explore and talk to experts about Microsoft Teams.

    Tuesday, November 6, 2018 6:30 AM
  • Hello,

    no, no next hop is pointing to the same server for either one of those pools I have been talking about. The only difference is that the SfB pool doesnt have SIP federation next hop defined at all on Site level (but I used existing edge as next hop on frontend pool level - but that was not edge from that site where the problem is).

    Do you think defining SIP federation in Site federation route assignment will help?

    Thank you,

    Tomas

    Wednesday, November 7, 2018 10:05 PM
  • Hello,

    no, no next hop is pointing to the same server for either one of those pools I have been talking about. The only difference is that the SfB pool doesnt have SIP federation next hop defined at all on Site level (but I used existing edge as next hop on frontend pool level - but that was not edge from that site where the problem is).

    Do you think defining SIP federation in Site federation route assignment will help?

    Thank you,

    Tomas

    Hello,

    any news regarding this?

    Thank you,

    Tomas

    Monday, November 12, 2018 8:41 AM
  • Hey Hi !

    The Skype for Business client is designed to contact all the available pChat pools in the topology, even though there is no next hop pool configured.

    For Example, you have 2 pChat pools. When the client signs-in, it will get both the pChat pool information as part of in-band provisioning and it send the INVITE packet to both the pChat pools and expect 200 OK from each of the pools. In your case it is getting the 200 OK (Positive response) from one pChat pool and not from other one.

    Please ensure connectivity between all the Front-end pool servers and all the Pchat pools is in place. Refer Ports and Protocols to know ports requirement.

    Note : No need to worry on SIP federation and etc.


    Neeranjan Shettar (MyPage)

    [If a post helps to resolve your issue, please click the "Mark as Answer" of that post or click Answered "Vote as helpful" button of that post. By marking a post as Answered or Helpful, you help others find the answer faster. ]

    Tuesday, November 13, 2018 6:59 AM
  • Hello,

    I think I an interesting update (at least for me:) )

    So I have one topology, with multiple sites, and I have one Lync 2013 persistent chat server on site 1, and one SfB persistent chat pool on site 5, and as I said, users homed on sites without persistent chat (meaning they have the one from that Lync 2013 site 1 set as default) are seeing that "Chatroom functionality may be limited due to an outage" error in their clients.

    Functionality itself is not limited in any way, because that SfB persitent chat pool is not yet production ready, so its not hosting any rooms (except for one testing).

    Now, when I try to list pers chat rooms from that Lync 2013 site, I can see all of them (those on Lync 2013 and also that Sfb one that I created for testing), but when I try to list rooms on the SfB (new) site, I get ONLY that test room that is created on new SfB pool.

    So it seems I forgot to open something somewhere on firewall and it is blocking access to chatrooms between those two sites. Can you please advise me what port should I allow between those 2 networks? I have tried to look it up in documentation but so far I have everything opened between those sites.

    Thank you for any advice.

    Tomas

    Friday, November 16, 2018 8:22 PM
  • 5061 TCP (SIP Communication) : Between Frontend servers and Pchat servers.

    5041 TCP (MTLS) : Between Pchat servers and Frontend servers.

    The typical flow of Pchat will be like below,

    When the client is trying to contact Lync2013 Pchat servers :

    SFB client >> Lync2013 FE > Lync2013 PCHAT

    When the client is trying to contact SFB Pchat servers :

    SFB client >> Lync2013 FE  >> SFB2015 FE >> SFB PCHAT servers

    Now lets think of where exactly the connectivity is missing !!

    • Lync2013 FE to SFB2013 FE ??
    • SFB2015 FE to SFB PCHAT ??

    Can you also check if the SFB PCHAT endpoint is created properly ? run the below command to verify, you will get the endpoints list, you will have to verify which one for SFB PCHAT by looking into  "PersistentChatPoolFqdn" parameter

    Get-CsPersistentChatEndpoint

    In case you do not see endpoint for SFB pCHat, you can create one by running the below command

    New-CsPersistentChatEndpoint -SipAddress "sip:PC-2SFB@litwareinc.com" -PersistentChatPoolFqdn "<SFB PCHAT Pool FQDN>" -DisplayName "Persistent Chat Endpoint SFB"


    Neeranjan Shettar (MyPage)

    [If a post helps to resolve your issue, please click the "Mark as Answer" of that post or click Answered "Vote as helpful" button of that post. By marking a post as Answered or Helpful, you help others find the answer faster. ]

    Friday, November 16, 2018 10:05 PM
  • Hello,

    yes, get-cspersistentchatendpoint returns (among other results) endpoint for the new pool, which is Enabled.

    Regarding those ports - are those ports the same also when we are talking about communication between sites?

    so PCHAT Site 1 should be able to contact FE from Site 2 over 5041?

    Or, to be more specific - what ports should be opened between frontends of 2 sites if I want them to pass Persistent chat data to each other?

    Thank you,

    Tomas

    EDIT:

    Here I can see that also some port 8011 is important if there is a multiserver topology in place (which, in my case, is true):

    https://docs.microsoft.com/en-us/skypeforbusiness/plan-your-deployment/persistent-chat-server/topology

    Can you please refer to that? Should I open 8011 only between servers in a SfB pool, or also between Lync 2013 Pchat in one site and the SfB Pchat servers in the other site?

    • Edited by TomasCrha Monday, November 19, 2018 12:40 PM
    Monday, November 19, 2018 12:28 PM
  • Hey Hi !!

    Port 8011 is required only between pChat frontend servers within a particular pChat pool. so no need to open between sites..

    ====Let me draw the scenario here====

    You have Multiple sites like below

    Site1 : Lync 2013 FE Pool A  and Lync 2013 Persistent Chat Pool

    Site2 : Lync 2013 FE Pool B

    Site3 : Lync 2013 FE Pool C

    Site 4: SFB 2015 Pool A and SFB Persistent Chat Pool

    So now, clients from each site to communicate respective Pchat pools (Lync andSFB) :

    For Lync 2013 Persistent Chat Pool:

    • Site1 Client > Lync 2013 FE Pool A(Site1) > Lync 2013 Pchat pool
    • Site2 Client > Lync 2013 FE Pool B(Site2) > Lync 2013 FE Pool A(Site1) > Lync 2013 Pchat pool
    • Site3 Client > Lync 2013 FE Pool C(Site3) > Lync 2013 FE Pool A(Site1) > Lync 2013 Pchat pool
    • Site4 Client > SFB 2015 FE Pool A(Site4) > Lync 2013 FE Pool A(Site1) > Lync 2013 Pchat pool

    For SFB 2015 Persistent Chat Pool:

    • Site1 Client > Lync 2013 FE Pool A(Site1) > SFB 2015 FE Pool(Site4) > SFB 2015 Pchat Pool
    • Site2 Client > Lync 2013 FE Pool B(Site2) > SFB 2015 FE Pool(Site4) > SFB 2015 Pchat Pool
    • Site3 Client > Lync 2013 FE Pool C(Site3) > SFB 2015 FE Pool(Site4) > SFB 2015 Pchat Pool
    • Site4 Client > SFB 2015 FE Pool A(Site4) >  SFB 2015 Pchat Pool

    Now lets talk about ports :

    Port 5061 : This port will be used only between Front-ends pools from irrespective of Pools to all the Front-end pools in the environment

    Port 5041 : This port will be used only between Pchat Pool and it's next hop Front-end pool.

    ============================================================

    My advise here is,

    1. Open 5061 and 5041 Port between all Site's Front-end pool and both the Pchat pools (Lync and SFB) bi-direction.
    2. Port 1433 from Lync Pchat pool to SFB Pchat pool's SQL Backend server(SQL).
    3. Port 1433 from SFB Pchat pool to Lync 2013 Pchat pool's SQL Backend server(SQL).

    ==I know this advise is no where documented, but based on my experience this what we can do in the Multi Pchat Pool environment.==

    Note : Ensure the SFB Pchat Pool servers are up and running and it's connectivity with it's Backend SQL server is fine.


    Neeranjan Shettar (MyPage)

    [If a post helps to resolve your issue, please click the "Mark as Answer" of that post or click Answered"Vote as helpful" button of that post. By marking a post as Answered or Helpful, you help others find the answer faster. ]


    Monday, November 19, 2018 3:32 PM
  • Thank you, I will handle FW openings where needed according to what you wrote, and will tell you if it helped.

    BR, Tomas

    EDIT:

    One more question: If I remove Persistent Chat endpoint, will the client stop trying to contact it upon sign in ?

    Thank you,

    BR, Tomas

    • Edited by TomasCrha Friday, November 23, 2018 2:51 PM
    Tuesday, November 20, 2018 10:24 AM