07-09-2012 06:31 AM - edited 03-03-2019 06:40 AM
Hi,
We're using a CSS501 to load balance http requests over three servers running windows server 2003 and IIS 6.
On Saturday I attempted to switch over to three new servers running the app on Server 2008 and IIS 7.5
I mirrored the used the existing setup to create the new environment (pwgecashwww01,02,03 Owner ACMENewServers and group NewServer).
CSS11501# show running-config
!Generated on 07/09/2012 06:26:18
!Active version: sg0750004
configure
!*************************** GLOBAL ***************************
ip route 0.0.0.0 0.0.0.0 192.168.139.254 1
!************************** CIRCUIT **************************
circuit VLAN1
ip address 192.168.139.9 255.255.255.0
!************************** SERVICE **************************
service OFPTSTWWW02
ip address 192.168.139.17
protocol tcp
port 80
keepalive type http
active
service OFPTSTWWW03
ip address 192.168.139.18
protocol tcp
port 80
keepalive type http
active
service ecash-1
ip address 192.168.139.3
protocol tcp
port 80
keepalive type http
active
service ecash-2
ip address 192.168.139.4
protocol tcp
port 80
keepalive type http
active
service ecash-3
ip address 192.168.139.5
protocol tcp
port 80
keepalive type http
active
service pwgecashwww01
ip address 192.168.139.10
protocol tcp
port 80
keepalive type http
active
service pwgecashwww02
port 80
keepalive type http
protocol tcp
ip address 192.168.139.11
active
service pwgecashwww03
ip address 192.168.139.12
keepalive type http
port 80
protocol tcp
active
!*************************** OWNER ***************************
owner ACME
content traffic
add service ecash-1
add service ecash-2
add service ecash-3
vip address 192.168.139.8
protocol tcp
port 80
balance aca
active
owner ACMENewServers
content traffic
vip address 192.168.139.13
protocol tcp
port 80
balance aca
add service pwgecashwww01
add service pwgecashwww02
add service pwgecashwww03
active
owner ACMETestEnv
content traffic
vip address 192.168.139.16
protocol tcp
port 80
balance aca
add service OFPTSTWWW02
add service OFPTSTWWW03
active
!*************************** GROUP ***************************
group TestServers
vip address 192.168.139.16
add destination service OFPTSTWWW02
add destination service OFPTSTWWW03
active
group newserver
vip address 192.168.139.13
add destination service pwgecashwww01
add destination service pwgecashwww02
add destination service pwgecashwww03
active
group servers
add destination service ecash-1
add destination service ecash-2
add destination service ecash-3
vip address 192.168.139.8
active
CSS11501#
I can connect to all three webservers (pwgecashwww01,02,03) directly. I can ping them from the CSS11501:
CSS11501# ping 192.168.139.10
Pinging 192.168.139.10 1 time(s)...
Working(-) 1/1
100% Success.
CSS11501# ping 192.168.139.11
Pinging 192.168.139.11 1 time(s)...
Working(-) 1/1
100% Success.
CSS11501# ping 192.168.139.12
Pinging 192.168.139.12 1 time(s)...
Working(-) 1/1
100% Success.
but no matter what I try I can't get them to show as Alive.
Name: pwgecashwww01 Index: 7
Type: Local State: Down
Rule ( 192.168.139.10 TCP 80 )
Session Redundancy: Disabled
Redirect Domain:
Redirect String: (null)
Keepalive: (HTTP:HEAD: 5 3 5 )
Keepalive Encryption: Disabled
Last Clearing of Stats Counters: 07/03/2012 02:41:48
Mtu: 1500 State Transitions: 0
Total Local Connections: 0 Total Backup Connections: 0
Current Local Connections: 0 Current Backup Connections: 0
Total Connections: 0 Max Connections: 65534
Total Reused Conns: 0
Weight: 1 Load: 255
Weight Reporting: None
Name: pwgecashwww02 Index: 8
Type: Local State: Down
Rule ( 192.168.139.11 TCP 80 )
Session Redundancy: Disabled
Redirect Domain:
Redirect String: (null)
Keepalive: (HTTP:HEAD: 5 3 5 )
Keepalive Encryption: Disabled
Last Clearing of Stats Counters: 07/03/2012 02:42:41
Mtu: 1500 State Transitions: 0
Total Local Connections: 0 Total Backup Connections: 0
Current Local Connections: 0 Current Backup Connections: 0
Total Connections: 0 Max Connections: 65534
Total Reused Conns: 0
Weight: 1 Load: 255
Weight Reporting: None
Name: pwgecashwww03 Index: 9
Type: Local State: Down
Rule ( 192.168.139.12 TCP 80 )
Session Redundancy: Disabled
Redirect Domain:
Redirect String: (null)
Keepalive: (HTTP:HEAD: 5 3 5 )
Keepalive Encryption: Disabled
Last Clearing of Stats Counters: 07/03/2012 02:45:43
Mtu: 1500 State Transitions: 0
Total Local Connections: 0 Total Backup Connections: 0
Current Local Connections: 0 Current Backup Connections: 0
Total Connections: 0 Max Connections: 65534
Total Reused Conns: 0
Weight: 1 Load: 255
Weight Reporting: None
It's the same application, the only difference is the OS and IIS version.
Any ideas?
Thanks
Neal
07-10-2012 07:35 AM
Arranged downtime and gave the switch a reboot last night which appears to have solved the issue.
Yaaay!
07-18-2012 04:46 AM
I notice that you don't appear to specify a specific uri for your HTTP head requests to the servers.
This can sometimes cause issues if your web server doesn't have a default web page set. The CSS when it makes keepalive http request is looking for a 200 reply code, and if it gets any other code will mark the service as down.
What we do is put a basic text file called id.html in the root of our webservers. It contains some useful troubleshooting info as it says "I'm server XXXX listening on port YYYY". Then our service definition has the lines:
keepalive type http
keepalive uri "/id.html"
An added plus to this is that if you're ever troubleshooting, you can try and access the id.html i.e. http://www.yoursite.com/id.html, and the response will tell you the actual web server to which you were connected.
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide