Similar to this question, but I don't think it applies in my case.
I have two total endpoints that are both showing a status of "Degraded."
I read that this usually indicates that the monitor configuration is resulting in a response other than 200. However, testing with curl shows a 200 response for both endpoints. I initially had it set to probe the homepage (which also returned a 200), but added a dedicated 'probe' page just in case, which made no difference.
Here's my traffic manager configuration (classic portal, but that shouldn't matter):
Browser network inspector:
Here's my curl results:
Apparently if all endpoints are degraded, the traffic manager will still function, but I'm concerned that it may be affecting our geo-redundancy.
Thank you in advance for any help!
The -L
in your curl
invocation is for "follow redirects" which is throwing you off.
Traffic Manager won't follow HTTP redirects!
Because your endpoints now return a 301 Moved Permanently
they are being rendered Degraded.
$ curl -Ii http://posguys-east.cloudapp.net/probe.html
HTTP/1.1 301 Moved Permanently
Content-Length: 147
Content-Type: text/html; charset=utf-8
Location: https://posguys.com/probe.html
This is most probably because you have a HTTP → HTTPS redirect configured in your web.config.
The solution is simple, remove and re-add endpoints with https://
instead.
It's not that simple, looks like your application is responding with a 301
redirect on https://
too.
$ curl -kIiL https://posguys-east.cloudapp.net/probe.html
HTTP/1.1 301 Moved Permanently
Content-Length: 147
Content-Type: text/html; charset=utf-8
Location: https://posguys.com/probe.html
Server: Microsoft-IIS/8.5
HTTP/1.1 200 OK
Cache-Control: max-age=5184000
Content-Length: 149
Content-Type: text/html
You'll need to address that part. Add an exception for /probe.html
to whatever rewrite rule you're hitting there. Test with curl
, but no -L
this time.