Hi expert, I need help to configure my F5 LTM to redirect the connection to another host in the pool when one of the host/node in the pool have internal server error 500. How can i redirect the request to another host in F5 LTM when hit this internal error? Thanks

Hi, if you set up an HTTP monitor to check for an HTTP status code of 200, any pool member serving a 500 will not be selected for load balancing

here is the output: [root@f5:Active:Standalone] config curl -s -v 172.16.5.76 1> /dev/null * About to connect() to 172.16.5.76 port 80 (0) * Trying 172.16.5.76... connected * Connected to 172.16.5.76 (172.16.5.76) port 80 (0) > GET / HTTP/1.1 > User-Agent: curl/7.19.7 (x86_64-redhat-linux-gnu) libcurl/7.19.7 OpenSSL/1.0.1l zlib/1.2.3 libidn/1.18 > Host: 172.16.5.76 > Accept: */* > < HTTP/1.1 200 OK < Content-Type: text/html < Last-Modified: Thu, 13 Apr 2017 19:11:24 GMT < Accept-Ranges: bytes < ETag: "31cc9dbe89b4d21:0" < Server: Microsoft-IIS/10.0 < X-Powered-By: ASP.NET < Date: Wed, 15 Nov 2017 20:15:47 GMT < Content-Length: 703 < { [data not shown] * Connection 0 to host 172.16.5.76 left intact

That looks fine. Try just putting 200 OK in the receive string. Any issues please output the tmsh/cli monitor config:

If you want allow additional status codes 301 and 302, you could do it as below in the receive string. (200 OK|301|302)

Still failed. No clue what going wrong here. My monitor setting as below: Node setting Pool setting

redirect IIS 500 error to another host in pool

13 Replies

Lee_Sutcliffe
Nacreous
Nov 14, 2017
Hi, if you set up an HTTP monitor to check for an HTTP status code of 200, any pool member serving a 500 will not be selected for load balancing

Shamsul_Alam_34

Nimbostratus

Nov 15, 2017

here is the output:

[root@f5:Active:Standalone] config  curl -s -v  172.16.5.76 1> /dev/null
* About to connect() to 172.16.5.76 port 80 (0)
*   Trying 172.16.5.76... connected
* Connected to 172.16.5.76 (172.16.5.76) port 80 (0)
> GET / HTTP/1.1
> User-Agent: curl/7.19.7 (x86_64-redhat-linux-gnu) libcurl/7.19.7 OpenSSL/1.0.1l zlib/1.2.3 libidn/1.18
> Host: 172.16.5.76
> Accept: */*
>
< HTTP/1.1 200 OK
< Content-Type: text/html
< Last-Modified: Thu, 13 Apr 2017 19:11:24 GMT
< Accept-Ranges: bytes
< ETag: "31cc9dbe89b4d21:0"
< Server: Microsoft-IIS/10.0
< X-Powered-By: ASP.NET
< Date: Wed, 15 Nov 2017 20:15:47 GMT
< Content-Length: 703
<
{ [data not shown]
* Connection 0 to host 172.16.5.76 left intact

Lee_Sutcliffe
Nacreous
Nov 15, 2017
That looks fine. Try just putting
200 OK
in the receive string. Any issues please output the tmsh/cli monitor config:
Shamsul_Alam_34
Nimbostratus
Nov 16, 2017
Still failed. No clue what going wrong here.

My monitor setting as below:

Node setting

Pool setting
Lee_Sutcliffe
Nacreous
Nov 16, 2017
Could you change your node monitor to gateway_icmp and remove https_443 from the pool so you're only using the Check_header_200-OK monitor

Gurus
Altostratus
Nov 15, 2017
If you want allow additional status codes 301 and 302, you could do it as below in the receive string.

(200 OK|301|302)
JG
Cumulonimbus
Nov 17, 2017
You need specify the HTTP version in your send string if your backend does not support HTTP/0.9:

GET / HTTP/1.0\r\n\r\n

.
JG
Cumulonimbus
Nov 17, 2017
You can enable debugging for the monitor. See https://networkology.net/2014/11/11/debug-f5-monitor-response-from-the-server/.
KeesvandenBos
MVP
Nov 17, 2017
Create a http monitor with alias port 80
send string:
GET / HTTP/1.1\r\nHost:\r\nConnection: Close\r\n\r\n

Receive string:

HTTP/1.[01] [23]0[0-6]

And enable it on your pool (please remove the other monitors)

Cheers,

Kees