Team Foundation Server (TFS) -The underlying connection was closed

image

Having build and upgraded Team Foundation server from 2010 to 2013, SQL Sever and Sharpoint Foundation 2013, running on VMWare ESX, the team found very frequent and random like connection dropped problems, most frequent being:
  • connection that was expected to be kept alive was closed
  • The underlying connection was closed

This happened trying to check things in, or other source tracking related activities against Team Foundation Server.

The hunt was on the the cause, a week of investigation (on and off), heading off looking at the IIS settings, checking settings for WCF, looking for possible max connection limits, working pools and recycling logs etc.

Also investigated a post regarding network interface card (NIC) off loading and to switch it off on virtual boxes (I’m not convinced by that).

Solution

For others in a similar position, I present our solution, the static IP address that was assigned to the server had accidentally had a a printer assigned to it too. So whenever the user came in to work and put the printer on (not every day), we would experience these problems, for a bit, until the printer went off again. All due to the duplicate IP address.