cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
Announcements

459
Views
0
Helpful
2
Replies
mchakir
Beginner

DLSw redundancy problem

| |

****** ******** | |

* *--CPA1--*7206-1*--| ******** |

*IBM * ******** |--*7206-3*--|

* * | ******** |

*HOST* | |

* * ******** |--******** |

* *--CPA2--*7206-2*--| *7206-4*--|

****** ******** | ******** |

| |

7206-1 and 7206-2 manage the the connection to the IBM host with a CPA card eich other.

7206-3 and 7206-4 manage the connection between the central site and the remotes sites via frame-relay network, they are border peer and DLSw load balancing coded. also manage the connection between a LAN of the central site and with bridge-group to reach an IBM HOST via 7206-1 and 7206-2.

The following is the configurations:

7206-3:

dlsw local-peer peer-id 3.3.3.3 group 3 border keepalive 0 promiscuous

dlsw load-balance round-robin

dlsw remote-peer 0 tcp 1.1.1.1 timeout 90

dlsw remote-peer 0 tcp 2.2.2.2 timeout 90

dlsw cache-ignore-netbios-datagram

dlsw bridge-group 1

7206-4:

dlsw local-peer peer-id 4.4.4.4 group 3 border keepalive 0 promiscuous

dlsw load-balance round-robin

dlsw remote-peer 0 tcp 1.1.1.1 timeout 90

dlsw remote-peer 0 tcp 2.2.2.2 timeout 90

dlsw cache-ignore-netbios-datagram

dlsw bridge-group 1

The probleme is:

In the central site, i see a slowness in the connection, and it is random, sometime a connction is very fast, other time it is very slow, and it can take up to 1,5 minute to connect.

After all that, i coded a dlsw redundancy in the 7206-3 and 7206-4 FastEthernets, the problem still exist in the LAN, and it cause a trouble in some remotes sites connected with frame-relay, that loose a connection to HOST.

Thank You for your help.

2 REPLIES 2
k.korlyakov
Beginner

1. To make load balancing work you will have to add the following commands:

dlsw timer explorer-wait-time ...

source-bridge ring-group ...

Ring-group number must be the same on 7206-3 & 7206-4 - it prevents unnecessary explorer flooding.

2. If bridge-group 1 on 7206-3 & 7206-4 shares the same Ethernet domain you must use DLSw+ Ethernet Redundancy feature.

3. Border peers in the same group must be in full-mesh and all members in the group must peers with all border peers in this group. On 7206-3 & 7206-4 you will have to add the following commands:

(7206-3)# dlsw remote-peer 0 tcp 4.4.4.4 ...

(7206-4)# dlsw remote-peer 0 tcp 3.3.3.3 ...

Send me more info about your cases. I will try to help you. Detailed picture will be very helpfull.

Content for Community-Ad