topic Re: Maestro cluster member "down" after VSX command in Hyperscale Firewall (Maestro)

Maestro cluster member "down" after VSX command

Gombodorj — Mon, 28 Nov 2022 07:36:35 GMT

Hello, I have one MHO-140, one SMS, two 6700 appliances working in VSX mode in one SG.

I was trying to configure bond interface and I used "set vsx off". Right after that member #2 has gone "down" in cphaprob state. I can ping/ssh to 192.0.2.2 from 192.0.2.1.

[Expert@mn-dc1-r1c1-sec-fw.vsx1-ch01-02:0]# cphaprob state

Cluster Mode: HA Over LS

ID Unique Address Assigned Load State Name

1 192.0.2.1 100% ACTIVE(!) mn-dc1-r1c1-sec-fw.vsx1-ch01-01
2 (local) 192.0.2.2 0% DOWN mn-dc1-r1c1-sec-fw.vsx1-ch01-02

Active PNOTEs: VSX, pull_config

Last member state change event:
Event Code: CLUS-112000
State change: INIT -> DOWN
Reason for state change: USER DEFINED PNOTE
Event time: Mon Nov 28 09:32:42 2022

[Expert@mn-dc1-r1c1-sec-fw.vsx1-ch01-02:0]#

I didn't anything other than "set vsx on/off". What could be the issue?

Re: Maestro cluster member "down" after VSX command

_Val_ — Mon, 28 Nov 2022 11:44:19 GMT

It says "USER DEFINED PNOTE". Any chance you used admin down on one of the shoulders?

Re: Maestro cluster member "down" after VSX command

Gombodorj — Mon, 28 Nov 2022 13:49:17 GMT

No, I didn't use admin down command. In fact, I tried to use admin up command it didnt work as well.

Re: Maestro cluster member "down" after VSX command

_Val_ — Mon, 28 Nov 2022 14:04:29 GMT

Please post output of "cphaprob -l list" command here from the faulty member

Re: Maestro cluster member "down" after VSX command

Daniel_Szydelko — Mon, 28 Nov 2022 14:31:42 GMT

Hello,

In addition you can paste the output from some extra cmds - just to show overall state:

g_all cpinfo -y all

vsx stat -v
vsx stat -l

vs_all g_all asg stat vs

asg stat vs all

Daniel.

Re: Maestro cluster member "down" after VSX command

Gombodorj — Tue, 29 Nov 2022 04:58:01 GMT

Hi, I put the output of those commands in one file. Thanks for help

Re: Maestro cluster member "down" after VSX command

_Val_ — Tue, 29 Nov 2022 09:10:29 GMT

Hi, this is what I see:

Device Name: VSX
Registration number: 3
Timeout: none
Current state: problem
Time since last report: 0.2 sec
Problematic VSIDs: 1.

Device Name: pull_config
Registration number: 14
Timeout: none
Current state: problem
Time since last report: 56.4 sec

One of the VSs reports a problem and cannot pull config scripts.

Try opening that VS object in the management for editing and press OK. This should initiate a config push to the VS. Finally, If it does not help, please open a TAC request.

Also, did you run "vsx mode on" on that cluster member after configuring the bond? Also, you are supposed to configure bonds from MHO, why would you touch an appliance?

Re: Maestro cluster member "down" after VSX command

Gombodorj — Tue, 29 Nov 2022 09:14:16 GMT

I was trying to bond Eth1-5, eth1-6 uplink ports to a switch. On the MHO there are no uplink ports, it only shows downlink ports.

Re: Maestro cluster member "down" after VSX command

emmap — Tue, 29 Nov 2022 10:11:31 GMT

Bonds are configured at the security group, not the MHO.

I'm not aware of setting vsx off causing a pnote (were in in global mode when you ran that?) but if it did, it should resolve after setting it back on again. Worst case you may need to reboot the SGM.

Re: Maestro cluster member "down" after VSX command

Daniel_Szydelko — Tue, 29 Nov 2022 10:13:56 GMT

You can try from SGM 1_2:

vsenv 0

cpha_blade_config pull_config all 192.0.2.1

vsenv 1

cpha_blade_config pull_config all 192.0.2.1

then reboot SGM 1_2 and check it again.

Anyway I think it isn't proper order to configure Maestro VSX SG.

Daniel.