I have two G450 upgraded to 39.20.0 With Board Insertion Failure. I configure G450, Install AVP, Connect to AVP, scroll and accept the agreement. Once that is completed, the G450 returns the board insertion failure and not able to connect to AVP through the Gateway. As stated above, I tried AVP 7.0, 7.1.3, and several G450 firmware versions all with the same results. I have the AVP working good in another, (much older) G450 on same firmware release. I can't find how to resolve or prevent these errors.
Avaya sent the following info below, I wasn't able to get the workaround to work.
kill -9 `pidof watchd` && /etc/init.d/watchd start
[AVP-750] s8300 not recognized in gateway and applications Created: 10/Jan/19 Updated: 16/Apr/19
Status: In Progress
Project: Appliance Virtualization Platform
Component/s: AvayaVIB
Affects Version/s: 7.1.3.2
Fix Version/s: 8.1.0.0, 7.1.3.4
Type: Bug Priority: P2 - Give High Attention
Reporter:
Assignee:
Resolution: Unresolved Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified
Issue Links: Split
Child AVP-775
[SPLIT:8.1.0.0] s8300 not recognized ... Open
Child AVP-776
[SPLIT:7.1.3.4] s8300 not recognized ... Open
Severity: 3-Medium
Found By: External Customer (Worst) - Production
Confirmed Products: CP::Aura Core::Avaya Aura® Appliance Virtualization Platform-7.1.3.2 (P1602)
Problem Type: Breakage of Feature
Found In Release: 7.1.3.2
Target Release: 8.1.0.0, 7.1.3.4
Story Points: 5
Customer Ticket: 1-6O0B1AQ
Sprint: AVP/AVPU/US/ASDT/CMP Sprint 84, AVP/AVPU/US/ASDT/CMP Sprint 85, AVP/AVPU/US/ASDT/CMP Sprint 86
Customer Visible Symptom: --Replace this Note--
This field may be visible to external Avaya customers. When editing this field, be aware of any sensitive information.
--Replace this Note--
Minimum Config: --Replace this Note--
This field may be visible to external Avaya customers. When editing this field, be aware of any sensitive information.
--Replace this Note--
Description
Customer reported that in CM the "list config media-gateway x" command should incorrect data for S8300 D and E boards on AVP 7.1.3.2.0 and G450 firmware 39.12.0.
list configuration media-gateway 1
SYSTEM CONFIGURATION
Board Assigned Ports
Number Board Type Code Vintage u=unassigned t=tti p=psa
001V1 ICC MM S8300X HW00 FW000
001V2 DCP MM MM712AP HW04 FW015 01 u u u u u u u
001V3 ANA MM MM714AP HW05 FW073 01 u u u 05 u u u
001V9 MG-ANNOUNCEMENT VMM-ANN 01 02 03 04 05 06 07 08
09 10 11 12 13 14 15 16
And on the MG:
fresno-g430-001(super)# show mm
MEDIA MODULE DESCRIPTION: v1
-------------------------------------------------
Type : ICC
Description : ICC Media Module
Serial Number : UNKNOWN
HW Vintage : 0
HW Suffix : X
FW Version : 0
No. of ports : 2
Faults : No Fault Messages
I have this same issue in the lab on S8300 D and E on AVP 7.1.0, 7.1.2, 7.1.3 and 7.1.3.2 and on gateway firmware G450 39.12.0, G430 38.20.1, G450 on 38.21.1. I have tired on on other combinations.
Comments
Comment by [ 15/Feb/19 ]
Hi),
I checked in AVP 7.1.3.3 on S8300D card, card details as below.
~ # cat /proc/cajun
cajun_major=3
cajun_sc_irq=7
cajun_base=8160
cajun_slot=1
cajun_sc_count=1
cajun_bailout=80
cajun_fw_vintage=1
cajun_board_suffix=D
cajun_board_serial_number=11WZ04424064
in GW, G450 it is shown as,
solutionpune-SLS(super)# show mm 6
MEDIA MODULE DESCRIPTION: v6
-------------------------------------------------
Type : MM712
Description : DCP Media Module
Serial Number : 081615201628
HW Vintage : 7
HW Suffix : A
FW Version : 14
No. of ports : 8
Faults : No Fault Messages
This is an MGC controlled Media Module, check
the MGC for additional status information.
solutionpune-SLS(super)# show mm 1
MEDIA MODULE DESCRIPTION: v1
-------------------------------------------------
Type : ICC
Description : ICC Media Module
Serial Number : UNKNOWN
HW Vintage : 0
HW Suffix : X
FW Version : 0
No. of ports : 2
Faults : No Fault Messages
This is an MGC controlled Media Module, check
the MGC for additional status information.
May I know since which build, fields are displayed wrong. I do not have any history on it.
Comment by ) [ 06/Mar/19 ]
CPLD driver creates /proc/cajun file with hardware card details and watchd process reads these entries and sends it to Media gateway via heartbeat.
While booting up, watchd starts first and then CPLD drivers is loaded.
# grep -e "watchd " -e "cpld" syslog.log
2019-03-06T19:43:03Z root: init Running watchd start
2019-03-06T19:43:08Z root: init Running 90.cplduser.sh
2019-03-06T19:43:09Z cplduser: cplduser: starting vmk_MgmtUserBegin
watchd thread exits when /proc/cajun file is absent. To fix this behavior, init sequence has to be reordered. As VMware ESXi take cares of startup watchd on every reboot and we have no control of it. So to fix this, either we should have way in watchd to wait till CPLD initializes or we have to restart watchd again after CPLD initializes.
Workaround -
1. Kill watchd process and start it again,
kill -9 `pidof watchd` && /etc/init.d/watchd start