I woke up this morning to a notification from uptimerobot that a site was down. I checked it myself, and it was not.
I ssh'd in and checked the server logs, and while normal requests were being dealt with fine, uptimerobot's were looking like this:
1.2.3.4 - - [23/Oct/2015:06:41:06 +0000] "GET / HTTP/1.1" 499 0 "-" "Mozilla/5.0+(compatible; UptimeRobot/2.0; http://www.uptimerobot.com/)"
1.2.3.4 - - [23/Oct/2015:06:41:36 +0000] "GET / HTTP/1.1" 301 5 "-" "Mozilla/5.0+(compatible; UptimeRobot/2.0; http://www.uptimerobot.com/)"
1.2.3.4 - - [23/Oct/2015:06:42:06 +0000] "GET / HTTP/1.1" 499 0 "-" "Mozilla/5.0+(compatible; UptimeRobot/2.0; http://www.uptimerobot.com/)"
1.2.3.4 - - [23/Oct/2015:06:42:36 +0000] "GET / HTTP/1.1" 301 5 "-" "Mozilla/5.0+(compatible; UptimeRobot/2.0; http://www.uptimerobot.com/)"
1.2.3.4 - - [23/Oct/2015:06:43:07 +0000] "GET / HTTP/1.1" 499 0 "-" "Mozilla/5.0+(compatible; UptimeRobot/2.0; http://www.uptimerobot.com/)"
499 is an nginx response:
Th 301's are redirects from https to http that uptimerobot is unable to remember (I've updated uptimerobot's settings now, the problem persists). For reference, here is what the log looked like an hour ago:
1.2.3.4 - - [23/Oct/2015:05:43:07 +0000] "HEAD / HTTP/1.1" 301 0 "-" "Mozilla/5.0+(compatible; UptimeRobot/2.0; http://www.uptimerobot.com/)"
1.2.3.4 - - [23/Oct/2015:05:43:09 +0000] "GET / HTTP/1.1" 200 20564 "-" "Mozilla/5.0+(compatible; UptimeRobot/2.0; http://www.uptimerobot.com/)"
1.2.3.4 - - [23/Oct/2015:05:48:07 +0000] "HEAD / HTTP/1.1" 301 0 "-" "Mozilla/5.0+(compatible; UptimeRobot/2.0; http://www.uptimerobot.com/)"
1.2.3.4 - - [23/Oct/2015:05:48:09 +0000] "GET / HTTP/1.1" 200 20564 "-" "Mozilla/5.0+(compatible; UptimeRobot/2.0; http://www.uptimerobot.com/)"
1.2.3.4 - - [23/Oct/2015:05:53:07 +0000] "HEAD / HTTP/1.1" 301 0 "-" "Mozilla/5.0+(compatible; UptimeRobot/2.0; http://www.uptimerobot.com/)"
1.2.3.4 - - [23/Oct/2015:05:53:09 +0000] "GET / HTTP/1.1" 200 20564 "-" "Mozilla/5.0+(compatible; UptimeRobot/2.0; http://www.uptimerobot.com/)"
Nobody has touched the server in that time. The requests (that return 499) are being routed through to the underlying server (django) ok.
Is this something wrong on my server, or is this uptimerobot's problem? It's not happening for any other sites I'm monitoring with them, which have the same setups.
0 Answers