cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
932
Views
0
Helpful
2
Replies

Does the nV Satellite Dual Host/Home support preemption?

Ahmed Muhi
Level 1
Level 1

Hello All,

This is a follow-up discussion to a previous one that I have opened previously, I have an NCS5002 acting as a Satellite, and two ASRs 9006/9910 acting as Hosts using IOS-XR 6.1.2, in an Active/Standby nV Satellite scenario. below are the configs I am using:

ASR9910

RP/0/RSP0/CPU0:A99k-Lab#sho run nv
Fri Feb 17 07:11:13.199 UTC
nv
satellite 100
type ncs5002
redundancy
host-priority 10
!
serial-number FOC1946R0D2
!
!

RP/0/RSP0/CPU0:A99k-Lab#sho run redundancy
Fri Feb 17 07:11:16.826 UTC
redundancy
iccp
group 1
member
neighbor 10.0.0.2
!
backbone
interface TenGigE0/4/0/35
!
!
!
!

RP/0/RSP0/CPU0:A99k-Lab#

ASR9006

RP/0/RSP0/CPU0:ASR9006-Lab#sho run nv
Sat Feb 18 14:05:27.961 UTC
nv
satellite 100
type ncs5002
redundancy
host-priority 20
!
serial-number FOC1946R0D2
!
!

RP/0/RSP0/CPU0:ASR9006-Lab#sho run redundancy
Sat Feb 18 14:05:30.338 UTC
redundancy
iccp
group 1
member
neighbor 10.0.0.1
!
backbone
interface TenGigE0/1/0/29
!
!
!
!

RP/0/RSP0/CPU0:ASR9006-Lab#

Now my problem is when I shutdown the ICL between the Active Host "the ASR 9910" and the Satellite, the Standby Host "the ASR 9006" would take over immediately and becomes the Active Host, which is great, still when I no shut the ICL link between the "Old Active Host with the better/Numerically-Lower host priority 10 compared to 20 on the Standby Host" the link goes to this:

RP/0/RSP0/CPU0:A99k-Lab#sho int desc | i TO-NCS
Fri Feb 17 07:23:28.285 UTC
Te0/4/0/0 up down TO-NCS5002-Lab
RP/0/RSP0/CPU0:A99k-Lab#

and the A99K would not take over even after the delayed switchback timer would expire " Five minutes according to this document I waited ten minutes". Also please check the below output:

RP/0/RSP0/CPU0:A99k-Lab#sho nv satellite status
Fri Feb 17 07:27:38.063 UTC
Satellite 100
-------------
Status: Discovery Stalled; Conflict: no Identification received yet
Type: ncs5002
Displayed device name: Sat100
IPv4 address: 10.0.100.1 (auto, VRF: **nVSatellite)
Serial Number: FOC1946R0D2
Received candidate fabric ports:
nVFabric-TenGigE0/0/78-79 (permanent)
nVFabric-HundredGigE0/1/0-3 (permanent)
Configured satellite fabric links:
TenGigE0/4/0/0
--------------
Status: Discovery Stalled; Conflict: no Identification received yet
Remote ports: TenGigE0/0/0-77,79
Rejected remote ports: TenGigE0/0/78
RP/0/RSP0/CPU0:A99k-Lab#

RP/0/RSP0/CPU0:A99k-Lab#sho nv satellite redundancy
Fri Feb 17 07:27:56.636 UTC
Satellite 100:
Local Priority: 0x80480a0000ff0000
Satellite not reachable
Delaying switchback
Configured priority: 10
Remote Priority: 0x0000140000010001
Configured priority: 20
Number of hops: 1

RP/0/RSP0/CPU0:A99k-Lab#

RP/0/RSP0/CPU0:ASR9006-Lab#sho nv satellite status
Sat Feb 18 14:21:52.159 UTC
Satellite 100
-------------
Status: Connected (Stable)
Redundancy: Active (Group: 1)
Type: ncs5002
Displayed device name: Sat100
MAC address: c472.95a7.a325
IPv4 address: 10.0.100.1 (auto, VRF: **nVSatellite)
Serial Number: FOC1946R0D2
Remote version: Compatible (latest version)
IOFPGA: 0.16
MB_MIFPGA: 0.16
DB_MIFPGA: 0.16
BIOS: 1.08
XR: 6.1.2 (Latest)
Received candidate fabric ports:
nVFabric-TenGigE0/0/78-79 (permanent)
nVFabric-HundredGigE0/1/0-3 (permanent)
Configured satellite fabric links:
TenGigE0/1/0/0
--------------
Status: Satellite Ready
Remote ports: TenGigE0/0/0-78
Rejected remote ports: TenGigE0/0/79
RP/0/RSP0/CPU0:ASR9006-Lab#

RP/0/RSP0/CPU0:ASR9006-Lab#sho nv satellite redundancy
Sat Feb 18 14:22:13.781 UTC
Satellite 100:
Local Priority: 0x0000140000010001
Configured priority: 20
Number of hops: 1
Remote Priority: 0x80480a0000ff0000
Satellite not reachable
Delaying switchback
Configured priority: 10

RP/0/RSP0/CPU0:ASR9006-Lab#

The funny thing though is that when I shutdown the ICL link between the ASR9006 "the Active Host with the worse priority" and the Satellite, the ASR9910 would take over immediately.

I have played over with the above scenario, shutting down, the ICL link between the Active Host and the Satellite, and each time the Standby Host would always take over no matter what the priority was, still preemption would not work.

Thank you so much for your time and kind support.

Kind regards,

Ahmed Muhi

1 Accepted Solution

Accepted Solutions

xthuijs
Cisco Employee
Cisco Employee

hi ahmed,

yeah preemption like hsrp routers have isnot there with the satellite dual homing.

you would be able to somewhat automate this via EEM scripting:

trigger on a syslog icl linkupevent and configure a manual/sw switchover.

xander

View solution in original post

2 Replies 2

xthuijs
Cisco Employee
Cisco Employee

hi ahmed,

yeah preemption like hsrp routers have isnot there with the satellite dual homing.

you would be able to somewhat automate this via EEM scripting:

trigger on a syslog icl linkupevent and configure a manual/sw switchover.

xander

Hello Xander,

Thank you so much for your kind reply, I have removed the configuration and re-created the scenario from scratch, and now it is working. I have collected the below results, maybe preemption was introduced in IOS-XR 6.1.2 .

I have simulated the Active host failure by shutting down the link to the Satellite, and the Standby Host took over, after re-enabling the port again the Active host came back up and I got the following result:

RP/0/RSP0/CPU0:A99k-Lab#show nv satellite status
Sun Mar 5 00:44:06.553 UTC
Satellite 100
-------------
Status: Connected (Stable)
Redundancy: Standby (Group: 1) (Recovery Delay remaining: 4m 50s)
Type: ncs5002
Displayed device name: Sat100
MAC address: c472.95a7.a325
IPv4 address: 10.0.100.1 (auto, VRF: **nVSatellite)
Serial Number: FOC1946R0D2
Remote version: Compatible (latest version)
IOFPGA: 0.16
MB_MIFPGA: 0.16
DB_MIFPGA: 0.16
BIOS: 1.08
XR: 6.1.2 (Latest)
Received candidate fabric ports:
nVFabric-TenGigE0/0/78-79 (permanent)
nVFabric-HundredGigE0/1/0-3 (permanent)
Configured satellite fabric links:
TenGigE0/4/0/0
--------------
Status: Satellite Ready
Remote ports: TenGigE0/0/0-77
RP/0/RSP0/CPU0:A99k-Lab#

RP/0/RSP0/CPU0:A99k-Lab#sho nv satellite redundancy
Sun Mar 5 00:44:34.154 UTC
Satellite 100:
Local Priority: 0x00480a0000010000
Delaying switchback
Configured priority: 10
Number of hops: 1
Remote Priority: 0x0008140000010001
Configured priority: 20
Number of hops: 1

RP/0/RSP0/CPU0:A99k-Lab#

When the recovery delay expired, the Host (A99K) with the better priority took over and I got the follow Syslog messages:

RP/0/RSP0/CPU0:A99k-Lab#
Mar 5 00:48:53.922 : icpe_satmgr[1191]: %PKT_INFRA-ICPE_GCO-5-SATELLITE_REDUNDANCY : This device is now the active device for Satellite 100.

RP/0/RSP0/CPU0:ASR9006-Lab#
RP/0/RSP0/CPU0:Mar 6 07:42:23.114 : icpe_satmgr[1197]: %PKT_INFRA-ICPE_GCO-5-SATELLITE_REDUNDANCY : This device is now the standby device for Satellite 100.

And now the A99K is the Active Host

RP/0/RSP0/CPU0:A99k-Lab#show nv satellite status
Sun Mar 5 00:54:28.737 UTC
Satellite 100
-------------
Status: Connected (Stable)
Redundancy: Active (Group: 1)
Type: ncs5002
Displayed device name: Sat100
MAC address: c472.95a7.a325
IPv4 address: 10.0.100.1 (auto, VRF: **nVSatellite)
Serial Number: FOC1946R0D2
Remote version: Compatible (latest version)
IOFPGA: 0.16
MB_MIFPGA: 0.16
DB_MIFPGA: 0.16
BIOS: 1.08
XR: 6.1.2 (Latest)
Received candidate fabric ports:
nVFabric-TenGigE0/0/78-79 (permanent)
nVFabric-HundredGigE0/1/0-3 (permanent)
Configured satellite fabric links:
TenGigE0/4/0/0
--------------
Status: Satellite Ready
Remote ports: TenGigE0/0/0-77
RP/0/RSP0/CPU0:A99k-Lab#