UCS Monitoring Solution

[jchenard]

I have been searching for an easy to use solution to monitor my UCS Infrastructure. We have various monitoring and alerting platforms but all seem to fall short. We simply want to have a platform that monitors UCS at an infrastructure level - performance, availability, reporting on firmware levels and such.

Below is a list of what we currently use:

• CA Nimsoft - This system monitors the server components but does not give us visibility to UCS at a logical level. Also no inventory items like serial numbers.
• Solarwinds - We have had limited success with Solarwinds. It does a decent job with basic monitoring but does not go very deep on the UCS platform. It is also very bulky on the reporting side. How can I get a report with all serial numbers and firmware with this product?
• UCS Performance Monitor - We have this running but do not really find any value in it.

We are also looking at the following:

• ScienceLogic - Expensive, but looks to give us what we need and much more. Will require heavy resources to stand this up internally.
• UCSHealth.com - Currently testing this SaaS offering. I am impressed with how quickly the data was available once we had the collector installed. Pricing is per server per month. Anyone else have experience with this solution?

Am I missing other options?

John

[brihopki]

One option might be to use the UCSMSDK to create your own script/solution for this.  You could easily setup something that could poll your UCSM domain and get out any information you want.

Documentation:

https://communities.cisco.com/docs/DOC-64378

Let me know if you have questions regarding it, but you can get information like firmware, etc by doing this and you can do it for multiple UCS domains.

[jchenard]

I use the UCS Powertool and scripts all the time and have automated many tasks but would rather not reinvent the wheel if there is a solution which provides just what I need.

The ucshealth.com solution seems to be working well for inventory and data exports and a pretty unique knowledge base lined to error IDs.

Thank you for your reply

[brihopki]

I'm sure they are doing the same thing, so you could build your own internally or pay someone to do it for you.

I will have some sample code up soon on my github to basically build your own type of system like this.  I have already created a mock up API that you send ucsm login information to and get back firmware, faults, etc.

Thanks