• October 31, 2020, 05:42:12 AM
Welcome, Guest. Please login or register. Registration is free.
Did you miss your activation email?

Author Topic: RFS4000 Cluster Issue  (Read 4519 times)

0 Members and 1 Guest are viewing this topic.

Offline mattymo

  • Rookie
  • **
  • Posts: 4
RFS4000 Cluster Issue
« on: September 13, 2013, 09:40:15 AM »
Hi,

We have 2 rfs4000 wireless controllers. Both controllers are clustered and can see eachother. However, every time they try to update a socket error occurs and it seems not all the profiles are clustered correctly. One is set to master and the other is of course the slave. The error i keep getting is:

Cluster master config update to 01.77.06.EA failed, Err: cs_request(01.77.06.EA) socket error: (145, 'connect(((24577770, 29),))

It seems it cant update the slave. Are the certain ports that need to be opened for clustering? Both controllers have no issues finding eachother via traceroute/ping.

Thanks,

Matt


Offline noobie

  • Full Member
  • ***
  • Posts: 92
Re: RFS4000 Cluster Issue
« Reply #1 on: September 18, 2013, 01:55:21 PM »
Hi Matt,

cluster is formed using mint links. I am not sure about your exact config, but most probably you are using level 1 IP mint links between the cluster members. MINT is using UDP ports 24576 and 24577, you need to make sure that they are not blocked by any device on the way.
Check if you can see the other RFS via mint:
#show mint neighbors
You can also check the MINT ID of both controllers and see how stable mint ping is. Example:

#show mint id

#mint pint XX.XX.XX.XX (01.77.06.EA in your case)

How exactly those RFSes are connected (i.e. what is between them and which vendor it belongs to)?

Thanks.

Offline mattymo

  • Rookie
  • **
  • Posts: 4
Re: RFS4000 Cluster Issue
« Reply #2 on: September 21, 2013, 05:24:35 PM »
Thanks for the response!

I did a show mint id and it has no problem seeing the other controller. mint pinging works fine. Low MS and doesn't seem to be any visible. Its a brocade RFS4000 its connected directly to a Cisco 3550 switch and the switch is connected to a watchguard firewall. The controllers are at different campuses across a MPLS WAN link. Any other ideas?

Here are the logs:

Controller 1:

br-rfs4000-7706EA*#mint ping 01.77.06.D6
MiNT ping 01.77.06.D6 with 64 bytes of data.
 Response from 01.77.06.D6: id=1 time=6.497 ms
 Response from 01.77.06.D6: id=2 time=6.627 ms
 Response from 01.77.06.D6: id=3 time=6.502 ms

--- 01.77.06.D6 ping statistics ---
3 packets transmitted, 3 packets received, 0% packet loss
round-trip min/avg/max = 6.497/6.542/6.627 ms
br-rfs4000-7706EA*#show mint neighbors
14 mint neighbors of 01.77.06.EA:
01.77.06.D6 (br-rfs4000-7706D6) at level 1, best adjacency ip-10.3.110.3:24576
01.9E.FC.BC (br650-9EFCBC-STEELES) at level 1, best adjacency vlan-1
01.9E.FC.FC (br650-9EFCFC-STEELES) at level 1, 2 adjacencies, 2 UP adjacencies, best adjacency vlan-1
01.9F.37.A0 (br650-9F37A0-FDR2) at level 1, best adjacency ip-10.2.98.11:24576
01.9F.37.D0 (br650-9F37D0-STEELES) at level 1, best adjacency vlan-1
01.9F.39.A4 (br650-9F39A4-STEELES) at level 1, best adjacency vlan-1
01.9F.3A.14 (br650-9F3A14-STEELES) at level 1, best adjacency vlan-1
01.9F.3A.54 (br650-9F3A54-STEELES) at level 1, best adjacency vlan-1
01.9F.3A.60 (br650-9F3A60-STEELES) at level 1, best adjacency vlan-1
01.9F.3A.68 (br650-9F3A68-STEELES) at level 1, best adjacency vlan-1
01.9F.3A.78 (br650-9F3A78-FDR1) at level 1, best adjacency ip-10.2.98.12:24576
01.9F.3A.7C (br650-9F3A7C-STEELES) at level 1, best adjacency vlan-1
01.9F.3A.8C (br650-9F3A8C-STEELES) at level 1, best adjacency vlan-1
01.9F.3A.9C (br650-9F3A9C-FDR3) at level 1, best adjacency ip-10.2.98.10:24576
br-rfs4000-7706EA*#

Controller 2:

br-rfs4000-7706D6#mint ping 01.77.06.EA
MiNT ping 01.77.06.EA with 64 bytes of data.
 Response from 01.77.06.EA: id=1 time=6.533 ms
 Response from 01.77.06.EA: id=2 time=6.596 ms
 Response from 01.77.06.EA: id=3 time=6.523 ms

--- 01.77.06.EA ping statistics ---
3 packets transmitted, 3 packets received, 0% packet loss
round-trip min/avg/max = 6.523/6.551/6.596 ms
br-rfs4000-7706D6#show mint neighbors
11 mint neighbors of 01.77.06.D6:
01.77.06.EA (br-rfs4000-7706EA) at level 1, best adjacency ip-10.1.110.26:24576
01.9E.FB.F4 (br650-9EFBF4-DUN1) at level 1, best adjacency ip-10.4.98.204:24576
01.9E.FC.9C (br650-9EFC9C-DAV-AP1) at level 1, 2 adjacencies, 2 UP adjacencies, best adjacency vlan-1
01.9E.FC.BC (br650-9EFCBC-STEELES) at level 1, best adjacency ip-10.1.110.53:24576
01.9E.FC.C0 (br650-9EFCC0-DUN3) at level 1, best adjacency ip-10.4.98.43:24576
01.9E.FC.C8 (br650-9EFCC8-55DUN-AP2) at level 1, best adjacency ip-10.4.55.102:24576
01.9E.FC.F8 (br650-9EFCF8-DUN2) at level 1, best adjacency ip-10.4.98.205:24576
01.9E.FD.14 (br650-9EFD14-55DUN-AP1) at level 1, best adjacency ip-10.4.55.101:24576
01.9E.FD.20 (br650-9EFD20-DAV-AP2) at level 1, 2 adjacencies, 2 UP adjacencies, best adjacency vlan-1
01.9E.FD.40 (br650-9EFD40-DAV-AP3) at level 1, 2 adjacencies, 2 UP adjacencies, best adjacency vlan-1
01.9F.37.D4 (br650-9F37D4-55DUN-AP3) at level 1, best adjacency ip-10.4.54.36:24576
« Last Edit: September 21, 2013, 05:27:30 PM by mattymo »

Offline noobie

  • Full Member
  • ***
  • Posts: 92
Re: RFS4000 Cluster Issue
« Reply #3 on: October 02, 2013, 11:21:25 AM »
Hi,

sorry for the late reply,
can you try mint ping with specifying packet size, default is 64, so specify 1500 (default in mint policy), see if this can get through MPLS link.

Since this is a cluster over WAN, I would suggest using level 2 mint links to form a cluster, but bear in mind that you would not be able to use tunneled WLANs in this case.

can you also share us the config here?
It can be that config from the master will result in a loss of connectivity and slave will reject it.

Thanks