Hello,
from some time now I'm experiencing error which is crashing whole system:
PAGE FAULT IN NONPAGED AREA - What failed: HTTP.sys
One of my client complained that his website is not accessible very often so I've suggested that I move his website to one of my other IIS servers. Mentioned problems started occuring after upgrading Windows Server 2012 R2 to 2016. So IMHO this is somehow related to SSL traffic. So I've started moving client website to another system and for various checks I've used tools.pingdom.com website where I've noticed that his website is using 32MB for every page load (thumbs which are not thumbs but big images) but ok this is his website and I just mentioned that to him. Weird thing happened later. System where I moved client website crashed the same way like system which was holding his website before (previous system was crashing on daily basis so I had to turn off kernel dump to make it start faster and I was considering moving all of clients websites to other systems). New system never crashed like today so I think that this is to big coincidence and this has to be related to his website. Bottom line my question is:
Does website which is accessible only (redirect to https from http) over SSL which loads 32MB od data (30.6MB images) can cause system crash in HTTP.sys module on daily basis? Everything points to this direction. I've asked client to contact his programmer and update his app to use thumbs or change pseudo thumbs to text links and I will see what will happen.
Thanks
EDIT: I've managed to contact clients programmer which fixed big images -> thumbs issue which reduced website content size loaded by client browsers from 32MB to 1.5MB. I will wait for few days and see if system crashes will stop to occur. If they will then this is a bug in IIS for sure. This might also be related to constant checking for website uptime by remote service (uptimerobot.com) but of course it does not justify system crash :).