Problems in a hosted Lync / SBS 2011 combination

Question

Hi,
The problem
My client is using hosted Lync from Startready / lync4all in combination with their local SBS 2011 server -on premise Exchange Server-. Especially multipoint conferencing -some internal nodes / externals on PSTN, is going bad all the time with a Lync 2013 client
as well as a Lync 2010 client. At no specific intervals internal nodes have to rejoin. Gibberish / dropped calls from the outside to nodes on the inside are happening on a regular basis.
Our suspects for causes are:
1. SBS 2011 Bad DNS settings on the SBS 2011 server like no correct forwarding to external dns´s, I´ll have to check that though, but I observed long times to connect to shares and wrong forwarding entries are a documented cause for long connection times on
SBS 2011. I tried the Lync Pre-Call diagnostic tool on Lync (2010) as well as W7 and W8 non lync enabled nodes, all failed to connect which points to the SBS 2011 DNS as a cause.
2. No STUN / edge traversal settings on clients
3. No IGMP snooping settings, no flow control, no DiffServ implemented on either clients or the Netgear FS728TP smart switch
4. No documentation of specific Lync settings on the Cisco ASA 5505 appliance are available, it is unsure what the supplier of the network did there. I suspect, among other things, that a session timer is screwing conference connectivity with the Lync server
of Startready.
5. No effort was made to effectuate Lync related settings on clients through Active Directory.
6. Only 10/100 Ethernet connections to clients are available
7. On the Netgear switch, turning on Flow Control, IGMP snooping and LLDP did not prevent Lync conference calls from being dropped.
Note: I tested the connection with the Lync servers of Startready more then ten times with and without many people using the network and Voice conversations. The connection to the Lync Servers is as flawless as a WAN connection can be. Response times average
at 11 ms with no packet loss at all, measured from an internal network endpoint. This excludes the Lync Servers of Startready as a cause and lets be fair, thousands of Startready customers are using this service every day without a glitch!
Any help with effectuating setting on clients through Active Diredtory methods is also welcome
Kind regards,
Tony Thijs

Answer 1

From my conclusion it should be a networ related issue.
Most Impact has the firewall, if statefull inspection is turned on for Lync traffic.
https 443 and stun 3478 will be use for the traffic to Lync online.
You can also check your Configuration with this website
http://trippams.online.lync.com/
You May also download the Lync 2013 precall diagnostic to test your performance
http://technet.microsoft.com/en-us/library/dn451255.aspx
regards Holger Technical Specialist UC