Asked by:
EWS unavailable; retrying connection, with some repeated users

Question
-
Hi
I have a weird problem with some users, the problem is intermittent and happens to random users but some of them are more likely to face EWS unavailable; retrying connection more frequent than others.
I have deleted the sip profile, the registry key of auto discovery under HKCU, then the problem was resolved for 2 or 3 days then reappeared again.
I have set the address book availability to web search only, re-deleted the sip profile and the registry key, then the problem was solved then reappeared again for the same users, some others report that they face this issue randomly and sometimes it is resolved without any intervention.
Meanwhile, during the problem they are able to browse to the EWS and receive the correct response.
I have ran out of ideas, thus any help will be much appreciated.
Thanks
Tuesday, October 25, 2016 4:10 AM
All replies
-
EWS information is cached in the registry with a default TTL value of 24 hours
In the event of Autodiscover failure, retries on a three-minute interval over a period of two hours.If Autodiscover fails after three successive attempts, use MAPI but continue to retry.If Autodiscovery succeeds, refresh cached values on an hourly basis.
How about these users , are they using different sip domains
SFB tries to find EWS performing these two basic queries:
http://autodiscover.<smtpdomain>/autodiscover/autodiscover.xml
_autodiscover._tcp.<smtpdomain>
Check the CAS certificate is including the SAN names for all SMTP and SIP domainsLinus || Please mark posts as answers/helpful if it answers your question.
Tuesday, October 25, 2016 11:06 AM -
The domain is domainA.com, and this is where machines joined and Exchange & SFB installed and records created.
While users have sip@domainB.com.
In this domain B:
We have:
_autodiscover
_sipinternaltls
lyncdiscoverinternal A record
mail A record
but no Autodiscover (which is in domain A)
All these records point to correct addresses and IPs of domain A
So any suggestions?
- Edited by Mohamed M. Aly Wednesday, October 26, 2016 4:55 AM
Wednesday, October 26, 2016 2:26 AM -
Hello Mohamed,
Since you have the sip and email id different you can go ahead with the comparison check policy
Set-CsClientPolicy -Identity name -DisableEmailComparisonCheck $true
More info here
https://lyncdude.com/2014/05/27/complete-guide-for-troubleshooting-lync-outlook-clients-integration/
Linus || Please mark posts as answers/helpful if it answers your question.
Wednesday, October 26, 2016 5:49 AM -
Hi Mohamed M. Aly,
Welcome to post in our forum.
For the DNS records, based on my understanding, you need point to the Lync FE server in domain A.
You could check the integration between SFB server and Exchange server, for details, please refer to
https://technet.microsoft.com/en-us/library/jj688098.aspx
For your Lync client, please make sure it has the latest update.
Hope this reply helpful to you.
Alice Wang
TechNet Community Support
Please remember to mark the replies as answers if they help and unmark them if they provide no help.
If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com.Wednesday, October 26, 2016 6:02 AM -
Already did that, with no success with same users.
Others are working normally.
Wednesday, October 26, 2016 6:30 AM -
The problem is intermittent and with some users only, if the records are not correct then this would affect all of the users.
Wednesday, October 26, 2016 6:36 AM -
Hi,
is both exchange and SFB on-premise?? Could you describe your environment abit more?
ARe users (specially EWS problem users) successfully able to login with SFB mobile client?
Are they all on same SFB client version?
Does the following work:
Conversation History
Missed Conversations
Missed Calls
Voice Mail Playback
High-Resolution Photos on contacts
Meeting tab - with meetings from calendar
Contact Information Presence based on Calendar Information
Unified Contact StoreAre they on same OWA mailboxpolicy?
Please try this guide for EWS troubleshooting:
https://support.microsoft.com/nb-no/kb/3120929
Please mark as helpful if you find my contribution useful or as an answer if it does answer your question. That will encourage me - and others - to take time out to help you. Thank you! Off2work
- Edited by Off2work Wednesday, October 26, 2016 6:51 AM Edit
Wednesday, October 26, 2016 6:49 AM -
Hi Mohammed M. Aly,
Thanks for your response.
Since the issue just appeared on some users, so we suggest you compare the user attributes in AD between affected accounts and normal accounts, if any difference, please try to modify it.
Hope this reply helpful to you.
Alice Wang
TechNet Community Support
Please remember to mark the replies as answers if they help and unmark them if they provide no help.
If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com.- Proposed as answer by Alice-Wang Monday, October 31, 2016 10:02 AM
- Marked as answer by Alice-Wang Sunday, November 6, 2016 3:50 AM
- Unmarked as answer by Mohamed M. Aly Monday, November 7, 2016 2:08 AM
Thursday, October 27, 2016 9:06 AM -
Hello Mohammed M. Aly,
Would suggest to first narrow down to users in corporate LAN which are having issues,
1. Use fiddler to check the web traffic while the problem users is launched along with netmon to check which DNS record clients resolves for EWS query.
2. Try the host entry for the Autodiscover record directly to exchange to isolate if its DNS causing the issue
3. Isolate users if in case affected users had recently changed their passwords for AD accounts
4. Delete the cache credentials on the problem machine from "control key mgr.dll" and purge
5. Switch the working and non-working users computer and check if the response for query changes
Thursday, October 27, 2016 3:59 PM