Pi 3.0 & 3.1 Licensing I'm confused.

[Richard Murphy]

Hi,

I am hoping someone (Marvin, I'm looking at you ) may be able to shed some light on something for me.

I have just read the release notes of Pi 3.1 and got very exited by some of the new features. Namely the below.

Data Center Assurance

The following enhancements have been made to Data Center:

• Data Center Topology—Allows you to view the physical topology of the Data Center devices. The Data Center topology map displays alarm badging for device icons to indicate the current alarm state of data center elements.
• Fabric Extender Visibility—Discovers the Fabric Extender (FEX) components that cover FEX inventory and FEX interfaces.
• Virtual Device Context 1 HOP Topology View—Includes the Virtual Device Context (VDC) 1- HOP topology view for Admin VDC. By launching the 360-degree view of Admin VDC from the Data Center topology, you can cross launch the respective managed VDC and VDC 1- HOP topology view.
• Network Virtual Port Channel Link Health Status—Allows you to view the virtual Port Channel (vPC) link health status when you choose Maps > Topology Maps > Datacenter Topology.
• Monitoring Nexus vPC Health—Introduces the Nexus virtual Port Channel (vPC) health monitoring policy. This feature fetches the configuration parameters from the operational Primary vPC-configured Nexus switch and looks for any discrepancies in the configuration, which can lead to inconsistencies by correlating with the secondary vPC-configured Nexus switch.
• Data Center REST APIs—Allows you to fetch cluster, data center, and Virtual Machine (VM) details from outside the Prime Infrastructure interface.

Now I have Pi 3.0 on a Gen2 appliance with 1100 Lifecycle license. At first I thought I would need to get Data Centre licenses. But then I read this in the Pi 3.1 admin guide.

Data Center: Regulates the number of blade servers being managed by UCS device(s) in Prime Infrastructure.The license count matches the number of blades or rack units associated with any UCS device.

Datacenter licenses are specific only to UCS devices (Blade count). Other Data Center devices, such as Nexus switches and MDS devices, are managed using a normal Lifecycle license.

So now I am very confused, if I upgrade to Pi 3.1 will I get the benefits for the Nexus switches mentioned in the release notes with my Lifecycle licences.

Thanks in advance.

Rich

[Marvin Rhoads]

Sorry Rich.

I have a couple of customers running 3.1 but none of them are managing FEXes or vPCs so I don't have any better information than what you've cited.

You might try sending an email with your question to ask-prime-infrastructure@cisco.com.

The business development team for PI monitors that alias and is pretty good about replying.

[Richard Murphy]

Thanks Marvin.

I think it may just be a case of do the upgrade and see if the features are included. If not then look at what license I may need to get.

I'll update this discussion once I've done it in case anyone else is as confused as me.

[Richard Murphy]

So upgrade one, looks like all I mention above is in the new version for me running old Lifecycle licenses.

However it now appears that all my Nexus are being talked to but on the Network Devices screen they all show Collection failure.... yet something else to sort out now.

[Marvin Rhoads · Accepted solution]

There's a bug with PI 3.1 and Nexus switches that could be responsible for the issue you mention.

https://bst.cloudapps.cisco.com/bugsearch/bug/CSCuw93479

I was seeing this at a customer site and just yesterday applied the new Device Pack 1 to their Prime Infrastructure. Although the release notes don't document that Device Pack 1 fixes the bug, it appears to have resolved it for me in this case.

http://www.cisco.com/c/en/us/support/cloud-systems-management/prime-infrastructure/products-release-notes-list.html

[Richard Murphy]

Thanks Marvin,

That was my next port of call.

[Richard Murphy]

Thanks Marvin

Device pack applied, however I had to delete and re-add the Nexus for them to complete inventory successfully.

Which of course has led to another problem where Prime wont let me delete some of the devices.

"10.x.x.x, Please check the following logs for failure reason: ifm_inventory.log,inventory.log, existenceInventory.log,persistence.log under /opt/CSCOlumos/logs Directory."

ifm_inventory.log has this in it.

"2016-07-08 11:42:01,454 [ThreadManager-BackgroundThreads-7] INFO  logging - IFM InventoryService info Error in deleting - 10.x.x.x: Could not execute JDBC batch update; SQL [delete from ProtocolEndpoint where ID=?]; constraint [WCSDBA.FK_VPCPEERLINK_LAGBASE]; nested exception is org.hibernate.exception.ConstraintViolationException: Could not execute JDBC batch update."

I feel a TAC case coming.

[Marvin Rhoads]

You're welcome Richard. Thanks for the rating.

Yes - your latest error message is most definitely TAC-worthy.

Once you get into SQL and JDBC errors, you are into developer territory and beyond what can be fixed in a forum thread.

[davidemoro]

Is this bux fixed by Cisco?