suddenly this morning I am having pure paths from every system profile occurring in incidents as time-outs. When I open the incident it is giving this in the state field. It says - The purepath was completed prematurely because the maximum GAT stall time was exceeded. On a side note, I am getting incident emails on this even though they are not set up.
The GAT is the global agent time and is a self-monitoring metric (this explains why you are getting alerts) which means that the system time gap between your agents and your DT server grew too large for the server to properly correlate methods to their transactions (PurePaths). You can actually chart this metric from the self-monitoring system profile.
Some preliminary things that you can check are in the server health dashboard. Please confirm that you are not seeing spikes in traffic that send your node per second count over the size specifications for your DT server. You can also check to with your network team to ensure there are no issues that popped up today.
This can be caused by many things and I would strongly suggest opening a support ticket to have them help with investigating it if none of those first steps lead to a resolution.