Exchange 2010 monitors

Hi, can you post the strings of the monitors that are failing?

 

thanks

Sure. I copied them right out of the config guide and change the hostnames to the vip hostnames and added a valid exchange login to the monitor. Here is one example

 

 

GET /owa/auth/logon.aspx?url=https://mail.example.com/owa/&reason=0 HTTP/1.1\r\

 

nUser-Agent: Mozilla/4.0\r\nHost: mail.example.com\r\n\r\n

 

 

This is the response I get from the cli

 

 

400 Bad Request

 

Content-Type: text/html; charset=us-ascii

 

Server: Microsoft-HTTPAPI/2.0

 

Date: Fri, 04 Nov 2011 18:57:51 GMT

 

Connection: close

 

Content-Length: 311

 

 

 

Bad Request

 

 

Bad Request

 

HTTP Error 400. The request is badly formed.

 

 

 

Connection closed by foreign host.

 

Try removing the second "\r\n" from the string. Our v11 monitor only uses one return/newline, so we may need to update the guide.

That didnt change anything. Here is another example GET /Microsoft-Server-ActiveSync/ HTTP/1.1\r\nHost: mail.example.com\r\n

 

The guide tells you to put Supported in the recive string but here is what I get back from the cli

 

 

HTTP/1.1 400 Bad Request

 

Content-Type: text/html; charset=us-ascii

 

Server: Microsoft-HTTPAPI/2.0

 

Date: Fri, 04 Nov 2011 19:25:22 GMT

 

Connection: close

 

Content-Length: 311

 

 

 

Bad Request

 

 

Bad Request

 

HTTP Error 400. The request is badly formed.

 

 

 

Connection closed by foreign host.

 

The send strings look good, so my first suspicion is a formatting problem. Just checking here, did you paste the send string into the monitor configuration as one single line? If it ends up on two lines it could fail.

 

 

Are any of the monitors working for you, or are they all failing?

Two questions: what version of F5 do you run? Also, did you enable SSL offload on the CAS?

I am running 10.2.2 and I did paste the strings into the gui and got a failure message that it wouldnt accept literal newlines so I adjusted accordingly. They are all failing excpet the monitor checking port 135 for the rpc vip. And the cas servers are doing the ssl encryption not the f5

A couple of things here:

 

 

1) Are you using monitors that are type HTTP or HTTPS? If the Client Access servers are expecting encrypted traffic, you'll need to make sure you're using HTTPS. This could be the cause of the problem you're seeing.

 

2) If you're sending traffic on to the Client Access servers encrypted, you still *have* to decrypt at the LTM, and then re-encrypt, in order to use any of the guidance we provide. Those virtuals all (well, except for MAPI) depend on being able to see the unencrypted traffic.For instance, in the HTTP traffic we need to see the URI path, the User-Agent, the Auth method and any cookies. In LTM parlance, you're going to need a clientssl profile and a serverssl profile.

When I put in a straight https monitor it fails but a straight http monitor works fine so I dont think the cas servers are expecting encrypted traffic. The VS has an clientssl profile that was created using the template but no server side ssl profile

Hi, subscribing to this thread hoping for solution. Due to time constraints, I gave up and just have the basic default HTTPS GET for functionality. Good luck!