A recent blog post from the Microsoft Exchange Team titled "Checklist for troubleshooting Outlook connectivity in Exchange 2013 and 2016 (on-premises)" promises to help us figure it out. There wasn't a check list, per se. Rather it was an organized collection of troubleshooting tips and techniques. We hope it helps us and helps you. (Note: this is NOT intended for Office 365 connectivity issues!)
First, here are examples of the connectivity problems they cite, and we've hit:
Here is a summary of their tips and recommendations:
* Ensure that everything is fully patched. We find Office Configuration Analyzer Tool (OffCAT) quite helpful. In fact, Microsoft Office released OffCat Version 2.2 June 2016. Download v2.2 here.
* They recommend cached mode vs. online mode to smooth out the user experience. We agree it helps, although it masks connectivity problems.
* Ensure CAS servers are not turning off NIC cards, use outdated drivers, or are not configured for power saving mode. The same holds true for the load balancer -- make sure keep-alive and idle timeouts are set above the 15 minute threshold.
* Too many cores: it's hard to believe that you can have too many cores, but you can. Don't have any more than 24 cores per server
* Configure Exchange performance monitoring ("perfwiz"). MS points you to two articles: Troubleshoot High CPU Utilization in Exchange 2013, and Exchange Monitoring tool, "Exmon"
* Logs: The article recommends Outlook logging, HTTP logging, IIS logging, Exchange Logging, and RPC logs. They recommend a tool, Log Parser Studio, to help parse the logs.
Writing this blog was the easy part. Now we'll have to try each suggestion until we discover what's causing the client connectivity problems.