I am asking this question to see if anyone else is seeing the same issue.

We have three DAG servers for Availability and the databases are setup as passive and lagged copies. When one Dag server goes down the database are mounted automatically to the passive copies, here very thing ok. But some clients whose mailboxes that are on the mounted databases on the downed DAG server start getting connection errors on the Outlook client  though the database are already mounted on another DAG server.Or at times the Outlook breaks completely and we have to delete the user profile and recreate it again to repair Outlook client.

Is this normal ? We never had this issue when we were on Exchange 2010. What could be the causes and could we avoid this ?

Could it be that since Outlook now uses https instead of tcp/ip connections that the clients are locked up on the cache trying to connect to the downed server ? If this is the Case is there away to avoid this ?

Activesync does not Show this problem.

Our servers are on Vmware.

I am asking this question to see if anyone else is seeing the same issue.

We have three DAG servers for Availability and the databases are setup as passive and lagged copies. When one Dag server goes down the database are mounted automatically to the passive copies, here very thing ok. But some clients whose mailboxes that are on the mounted databases on the downed DAG server start getting connection errors on the Outlook client  though the database are already mounted on another DAG server.Or at times the Outlook breaks completely and we have to delete the user profile and recreate it again to repair Outlook client.

Is this normal ? We never had this issue when we were on Exchange 2010. What could be the causes and could we avoid this ?

Could it be that since Outlook now uses https instead of tcp/ip connections that the clients are locked up on the cache trying to connect to the downed server ? If this is the Case is there away to avoid this ?

Activesync does not Show this problem.

Our servers are on Vmware.

What are you using for load balancing the client connections?

Were are using Citrix Netscaler Hardware Balancer, but for the last few days we bypassed the HLB and used round robin for some other reasons on our dns server.

10.x.x..x email.company.com

10.x..x.y email.company.com

10.x..x.z email.company.com

. So today when we got the problem the load balancer was bypassed. But may be one question, was that enough or we needed to add the autodiscover as well ?

i.e

10.x.x.x email.company.com autodiscovey.company.com

Were are using Citrix Netscaler Hardware Balancer, but for the last few days we bypassed the HLB and used round robin for some other reasons on our dns server.

10.x.x..x email.company.com

10.x..x.y email.company.com

10.x..x.z email.company.com

. So today when we got the problem the load balancer was bypassed. But may be one question, was that enough or we needed to add the autodiscover as well ?

i.e

10.x.x.x email.company.com autodiscovey.company.com

I would load balance all the URLS  that point to multiple servers under one namespace yes.

(AutoD, OAB, EWS, Outlook Anywhere etc..)

in otherwords this could be caused by the load balancer

in otherwords this could be caused by the load balancer

Indeed. Thats where I would start.

Thanks, let me how I can test this

I have bypassed all URLs and did  some tests, so after downing the DAG, the Outlook client showed "tyring to connect..." and took about 3 mins to reconnect, so is this normal ?

I have bypassed all URLs and did  some tests, so after downing the DAG, the Outlook client showed "tyring to connect..." and took about 3 mins to reconnect, so is this normal ?

Doesn't sound normal. I haven't used Citrix for a load balancer but you may want to open a case with them to ensure all the settings are correctly configured on the load balancer.

Ex: https://www.citrix.com/content/dam/citrix/en_us/documents/products-solutions/microsoft-exchange-2013-citrix-netscaler-deployment-guide.p

I have bypassed all the URL's and this morning one DAG went down and users got a disconnection for along period till we brought the server back

Ok the whole issue was that the DAG servers CPU was randomly shooting to %100 and that was why were experiencing this problem we the user connections.