I also suddenly got the "IOHandler value is invalid" message a few days ago on a HTTP sensor that had been working for ages, though not with the "Reset by peer" text you had.
In my case the webserver had an invalid SSL certificate, so I figured that was the reason and just deleted the sensor.
(see error message below)
Now I tried it again using the HTTP Full Web Page Sensor.
At first it just said "Sensor Timeout reached", I guess because the SSL certificate is fubar.
But when I changed the Browser Engine in PRTG to "Internet Explorer", it suddenly gave me an OK response.
So, the Browser Engine used by PRTG clearly has impact here.
Try monitoring using HTTP Full Web Page sensor using either "Chromium" or "Internet Explorer" as browser engine, in the Sensor Settings.
If you still get "Connection Reset by Peer", I suspect this is caused by the Exchange webserver.
I would install Wireshark on the PRTG server and monitor the HTTP traffic. (using a filter like "host a.b.c.d"
If you see the Exchange server sending a RST or FIN packet early on, this means the peer (Exchange server) has indeed reset the connection (RST=Reset).
Error message:
Failed to establish secure connection [Step 0] Error connecting with SSL. Error connecting with SSL. error:14082174:SSL routines:SSL3_CHECK_CERT_AND_ALGORITHM:dh key too small [Step 1] Error connecting with SSL. Error connecting with SSL. error:14082174:SSL routines:SSL3_CHECK_CERT_AND_ALGORITHM:dh key too small [Step 2] Error connecting with SSL. Error connecting with SSL. error:14082174:SSL routines:SSL3_CHECK_CERT_AND_ALGORITHM:dh key too small [Step 3] Error connecting with SSL. Error connecting with SSL. error:14082174:SSL routines:SSL3_CHECK_CERT_AND_ALGORITHM:dh key too small [Step 4] Error connecting with SSL. Error connecting with SSL. error:14077102:SSL routines:SSL23_GET_SERVER_HELLO:unsupported protocol [Unsecure] IOHandler value is not valid
Add comment