Connect APIC with Prime Infrastructure PNP

[steffen.bodensohn@heraeus.com]

Hello,

I want to connect APIC-EM with Prime Infrastructure.but it does not work.

Does anyone know if I have to consider something else  than inserting the APIC IP Address, Username, Password, etc. ?

I want to connect it to use APIC for Plug and Play Zero touch instead of Plug and Play Gateway.

Best regards,

Steffen

[yawming]

Please refer to the following document.

Use APIC-EM for PnP should work.

Solution Guide for Cisco Network Plug and Play - Cisco

Thanks,

Yawming

[maharbec]

We haved changed a last minute thing in APIC-EM API. Therefore we need PI 3.0.2 - the PI Patch will be out early december 2015.

It works CA2/3 with PI 3.0

Krg

Markus

[steffen.bodensohn@heraeus.com]

Hello Markus,

thanks for your reply.

I now installed CA2 release and PnP is working now.

But I´m getting a provisioning error:

Provisioning is done in VLAN1

LAN Management is done in VLAN 101

On the access switches I shut down Vlan1 in the provisioning configuration because I don´t need it anymore.

I think the problem is caused because provisioning is done within vlan1 and after I shut it down in the deployment configuration file APIC-EM is losing the connection.

How to solve this issue ?

DE-HAU-LAB-S201#sh pnp profile

Initiator Profile test: 1 open connections: 0 closing connections

Encap: pnp

WSSE header is not required. Configured authorization level is 1

Work-Request Tracking: Validation Yes, Violation 0, UDI , Correlator

PnP Response Pending: Retry Allowed 0, UDI , Correlator

Countdown: Security Unlock: S=3/F=0/T=0, Service Lock: S=0/F=0/T=0, Service Req Wait: S=1/F=0/T=15, Prxoy Req Wait S=75/F=0/T=0, Service Resp Ack: S=199/F=0/T=0

Max message (RX) is 50 Kbytes

XEP Faults are sent

Idle timeout infinite

Keepalive not configured

Reconnect time 60 seconds

Primary Transport:https to Host:172.18.37.122, Port:443, Src-Intf:-, VRF:-, URL pnp/WORK-REQUEST

Connected to the primary transport via https

Remote connection via HTTP client. URL https://172.18.37.122:443/pnp/WORK-REQUEST, post

Established at 10:28:51.957 UTC Fri Nov 20 2015

Tx 136234 bytes (380 msg), Tx 0 errors,

Last message sent at 12:04:14.273 UTC Fri Nov 20 2015

Rx 95222 bytes (380 msg), 0 empty msg

Last message received at 12:04:14.532 UTC Fri Nov 20 2015

Von: Markus Harbeck

Gesendet: Freitag, 20. November 2015 09:19

An: Bodensohn, Steffen <steffen.bodensohn@heraeus.com>

Betreff: Re: - Connect APIC with Prime Infrastructure PNP

Cisco Communities <https://communities.cisco.com/>

Connect APIC with Prime Infrastructure PNP

reply from Markus Harbeck<https://communities.cisco.com/people/maharbec> in Developer > Networking > SDN > APIC EM - View the full discussion<https://communities.cisco.com/message/198419#198419>

[maharbec]

Hi,

finally the Patch for PI 3.0 - (3.0.2) is available now. Check CCO.

YOu need to do some of the following: - the uplink goes into error.

1. use pnp vlan configuration on the uplink switch port

2. use no negotiate on the uplink

the uplink port goes into error becuase of SPT inconsistancy.

let me know if you need additonal help.

/Markus

[steffen.bodensohn@heraeus.com]

Hello Markus,

the uplink is working.

The initial deployment is done withing vlan1.

After the configuration is downloaded from APIC-EM:

• VLAN1 is shutdown
• Vlan101 management vlan is up

It seems that APIC-EM is waiting for a response from the pnp agent but is always receiving a request which forces the switch to load the configuration file over and over again.

This is the error message:

Received new work request from Agent while expecting work response. Retrying operation PROVISIONING_CONFIG

Any ideas ?

Is manual interaction allowed during deployment process ?

When deploying our configuration file a kron script triggers a tcl script to reload all interfaces automatically to ensure autosmartport is triggered. Could this cause the problem ?

Best regards,

Steffen

[maharbec]

Hi,

sorry i am on travel. So if the configuration is deploeyed we expect that APIC-EM can connect to the switch. I am not 100% sure i guess we connect / expect thre same IP on the device.

While booting no manual intervention after that, yes. But the deployment should go into provisioned if the entire configuration is written to the switch with success.

krg

Markus