Re: Prime Collaboration Installation

laffitte laurent · ‎12-09-2011

Hi,

I am currently trying to install an evaluation version of Cisco Prime Collaboration Manager on our Mock Up but I am confronted with an error when discovering the CTS Manager on the Prime Collaboration manager and I do not understand the error.

There is a valid Metrics Dashboard and Reporting API license in the CTS Manager as the LIC-CTS-MAN-RPT is installed and valid:

Name	Status
LIC-CTS-MAN-RPT	LICENSE_VALID

On the LDAP Server, I have created a group for the Prime Collaboration Manager as well as a user (primecollaboration manager) ; I have then added the user to the group, as shown in pictures '27 - LDAP - Prime Collab group', '27 - LDAP - Prime Collab user' and '28 - LDAP - Prime Collab user 1'.

On the CTSM, I have associated both the Live Desk Role and the Reporting API User role to the prime collaboration group (cf. picture '23 - CTSM access management - group roles added').

On the Prime Collaboration Manager, I have created a credential profile for the CTSM (cf. picture '26 - Prime Collab - discover devices 1').

However, when I click on the Verify button to check the profile I have created, the errors shown in pictures '26 - Prime Collab - discover devices error' and '26 - Prime Collab - discover devices error 1' appear.

I have looked for something related to such an error but did not managed to find anything.

I have tried to delete the CTSM in the current inventory but it tells that I cannot because is not in the unreachable state (indeed I can ping the CTSMan from the Prime Collab Manage in CLI).

Is there something I am doing wrong ?

laffitte laurent · ‎12-12-2011

Here are some logs I just downloaded from the Prime Collaboration Manager device.

IventoryDiscovery.log:

AccessLevelDiscovery.log:

12-Dec-2011|10:16:25.409|ERROR|AccessLevelDiscovery|pool-7-thread-45|getDeviceType(): SNMP get for sysOID and SysDesc has failed... Error while trying to run handler. Action : getsysObjectID, Handler : com.cisco.nm.pal.customhandler.snmp.SNMPTransportHandler. Error : Failed to perform SNMP Request: Received no response from 172.16.21.5/161. For SNMP GET request

12-Dec-2011|10:16:35.430|ERROR|AccessLevelDiscovery|pool-7-thread-45|getDeviceType(): SNMP get for sysOID and SysDesc has failed... Error while trying to run handler. Action : getsysObjectID, Handler : com.cisco.nm.pal.customhandler.snmp.SNMPTransportHandler. Error : Failed to perform SNMP Request: Received no response from 172.16.21.5/161. For SNMP GET request

Device.log:

I can see that the CTSM is considered as Inaccessible but I do not have a clue about the cause ; someone already had this type of issue when adding a CTSM to the Prime Collaboration Manager ?

Andrew Beezley · ‎12-12-2011

Hello Laurent - SNMP probing is the first form of validation that CPCM conducts during the Access Level discovery stage. If the device (in this case CTS Manager) fails to respond to SNMP (which it did), CPCM will try to match the other profiles. The Discovery is failing with state "INACCESSIBLE" which means none of the mandatory credentials were provided. So at a minimum, we know that the device is reachable on the network via ICMP since the state is not "UNREACHABLE".

From the logs below from AcessLevelDiscovery.log, you can see that the device was found reachable via ICMP so it moved to AL Discovery via SNMP which fails, it tries HTTP Probing, that fails and finally tries the Reporting API which also fails.

09-Dec-2011|10:01:22.513|INFO |AccessLevelDiscovery|http-443-2|:Device Credetail profile obtained for profileId:317 dc:id=317; name=CTSM; pattern=null; map={SNMP_READ_CS=(! removed), HTTP_USERNAME=admin, HTTP_PASSWORD=(! removed) ADDRESS_PATTERN=172.16.21.5, SNMP_VERSION=2c, SNMP_TIMEOUT=10000, SNMP_RETRIES=2}

.

, TestName:HTTP testResult:Failed reponseTime:5

, TestName:REPORTING API testResult:Failed reponseTime:0

]

The Device will need to move from INACCESSIBLE to ACCESSIBLE before Inventory Collection is started.

Things to check for are listed below:

- Is SNMP being blocked between CPCM and CTS Manager?

- Ensure SNMP version 2c is configured in UCM and not SNMP v3

- Is SNMP Process Enabled and running on CTS Manager?

=> from admin CLI: utils service list

=> from Web GUI: Configure => System Settings => SNMP

- Is HTTP being blocked between CPCM and CTS Manager?

By the way, what version CPCM and what version CTS-Manager are you deploying?

Kind regards,

-Andrew

laffitte laurent · ‎12-13-2011

Hi Andrew,

Thank you for your replay. I am using the last version of the PCM (1.1.0.10719) and the 1.8.0.0(582) ; I am using the evaluation version of the PCM, for now I do not have a license file.

I checked some of the things you mentioned in your previous post:

- SNMP version 2c was configured both on the CTSM and the PCM however, I had to restart the SNMP service on the CTSM for the communication to work properly,

- http is not blocked between the two devices, they are on the same network without any device between them.

Not knowing where to look to understand the HTTP issue, I used wireshark to make traces of the messages exchanged between the CTSM and the PCM: there was an authentication failure when PCM was trying to get connection on the CTSM.

I checked the user password on the LDAP server and the password configured on the associated PCM credentials but it did not change anything.

Finally, I tried the following:

- I placed the user dedicated to PCM created in LDAP in an administrator group,

- in the CTSM, I gave the administrator group the administrator role instead of the Live Desk role and the Reporting API role mentioned in the intall guide and it worked, the PCM got connected to the CTSM.

I assume this is not the normal behavior but I do not understand why I had to configure the PCM with such a user.

Also, do you have clues on the following states in the device inventory:

- Mediatrace Role: Unsupported

- IPSLA Role: Unsupported

- Performance Monitor: Unsupported

For the performance monitor to work, is there something to enable on the CTSM ?

Andrew Beezley · ‎12-13-2011

Hello Laurent - Glad to hear that you were able to resolve the communication failures between CPCM and CTS-Manager. That is interesting that you had to put the user created for CPCM into an Administrator role instead of the Live Desk and Reporting API role as designed. I will double-check on my end to see if that is / is not normal behavior and let you know. For now, it does not seem like normal behavior.

As far as Mediatrace, IPSLA and Performance Monitor are all Medianet capabilities that are not found/supported in all devices. CPCM will use Medianet to obtain jitter, packet loss, latency, DSCP values, etc... on those devices that support medianet technologies.

Below is a good link to read about Medianet and it capabilites and you will also find in release notes which devices support Medianet.

http://www.cisco.com/web/solutions/medianet/knowledgebase/index.html#~start

-Andrew

laffitte laurent · ‎12-13-2011

Hi Andrew,

it could be interesting to check wether it is a normal behavior or not yes

Thank you for the MEdianet link, I will read it asap to understand better the overall features.

Laurent