cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
452
Views
0
Helpful
2
Replies

round-robin problem

josephschung
Level 1
Level 1

I have two web servers (PISWEBP01 & PISWEBP02) with the following configuration on 11501. I take the default load balancing scheme (round-robin) and expect the connection will be distributed to the two server evenly.

But, the result is PISWEBP02 receive three time more connection than PISWEBP01. Any idea?

The web content is configured as below. Thanks.

content PIS-HTTP-VIP-1

vip address 10.106.13.224

redundant-index 33

add service PISWEBP01

add service PISWEBP02

protocol tcp

port 80

active

content PIS-SSL-VIP-1

vip address 10.106.13.224

redundant-index 36

advanced-balance ssl

application ssl

add service PISWEBP01

add service PISWEBP02

protocol tcp

port 443

url "/*"

active

2 Replies 2

Gilles Dufour
Cisco Employee
Cisco Employee

this is a sticky configuration.

If you had a server down for a while or started with one server than the other, it could explain that most sticky entries point to a single server.

You should try to remove stickyness and then re-apply it (see if that clears the sticky table).

Otherwise you will need to reboot.

Gilles.

Dear Gilles,

I modify the load balancing scheme to least connection (balance leastconn) for the content rule.

We have load test tool to generate loading and I execute "sh service XXX" comment to check the connection.

A snapshot is listed below:

=================

L1150102_WC# sh service PISWEBP01

Name: PISWEBP01 Index: 3

Type: Local State: Alive

Rule ( 10.106.13.20 ANY ANY )

Session Redundancy: Enabled

Redundancy Global Index: 31

Redirect Domain:

Redirect String:

Keepalive: (SSL-443 5 3 5 )

Keepalive Encryption: Disabled

Last Clearing of Stats Counters: 11/17/2008 22:57:37

Mtu: 1500 State Transitions: 202

Total Local Connections: 99272 Total Backup Connections: 0

Current Local Connections: 133 Current Backup Connections: 0

Total Connections: 99272 Max Connections: 65534

Total Reused Conns: 0

Weight: 1 Load: 14

Weight Reporting: None

L1150102_WC# sh service PISWEBP02

Name: PISWEBP02 Index: 4

Type: Local State: Alive

Rule ( 10.106.13.21 ANY ANY )

Session Redundancy: Enabled

Redundancy Global Index: 32

Redirect Domain:

Redirect String:

Keepalive: (SSL-443 5 3 5 )

Keepalive Encryption: Disabled

Last Clearing of Stats Counters: 11/17/2008 22:57:37

Mtu: 1500 State Transitions: 3538

Total Local Connections: 29568 Total Backup Connections: 0

Current Local Connections: 38 Current Backup Connections: 0

Total Connections: 29568 Max Connections: 65534

Total Reused Conns: 0

Weight: 1 Load: 2

Weight Reporting: None

=================

Now, the situation is reversed. The "Current Local Connection" of PISWEBP01 server now receive about three times more connections than PISWEBP02 server.

Any idea?

Thanks.

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: