Map service failures on load

Hi all.

Just wondering if anyone out there has any experience setting the Failure Timeout (seconds) value - as part of a Map Service definition/config. 

We, at present are getting some pretty random errors on load of the sites, that inform us one or more services have not loaded, and with our Failure Action set to Warn right now, the erroneous map service remains in the layer list but for all intents and purposes, it's dead in the water. Any click on the layer is not registered in the client!  

We've checked many times that all our services are running, across multiple AGS nodes, and 99% of the time, it all loads fine (I personally cannot reproduce the issue), so it's REALLY random. We will look at recycling the app pool every 12 hours to see if that makes any difference, but for starters, we are wondering if anyone knows anything about the Failure Timeout setting, and what the default setting is or what it actually does?

thanks as always

GF

PS: essentials and viewer version 4.14.2