5 Replies Latest reply: May 1, 2012 8:57 AM by Laurent Matheo RSS

Automated data population to CMDB

Lin Zhu

Hi,

 

I am new to BMC products. I am trying to find a best practice to automatically populate data to CMDB. The data could be some data from other databases, or flat files, spread sheets, xml etc. We are still in the process of collecting data (CIs and their relationships, data model won't be too complicated) and what we are sure right now is that the data will be really huge (almost 1 million CIs, several GBs, could be more. No way to cut down).

 

For some reason, federation and discovery tool will not be considered for this project. We have to populate large data into CMDB as required (it is a one-direction duplicating data with ongoing synchronization with the CMDB). I am reading a lot of BMC documents and trying to find out the best practice and required tools and technologies. Of couse I also read some discussions in this communities. My first impresson is either to use "Atrium Integrator" or we develop our own program through APIs. Am I correct?

 

"BMC recommends that Atrium Integrator should be used for all new data transfers because BMC Atrium Integration Engine will be deprecated in a future release." is what I find out from the Atrium Integrator7.6.04_UsersGuide.pdf. Can AtriumIntegrator achieve auto data population? How handy is it? Does it require a seperate license? What's the cost of the license?

 

Will it be cost-effective to build our own program to populate huge data into CMDB automatically (we have developers also)? Java APIs or Web services, which is recommended for large data population? Can it set normalization or reconciliation rules? Can it synchronize the source data to CMDB?

 

Atrium Integrator vs. APIs? Any pros and cons for large data population?

 

Thanks in advance! Any suggestion / recommendation will be appreciated.

  • 1. Re: Automated data population to CMDB
    Dirk Anderson

    Atrium Integrator is your best bet unless you have a slew of developers ready to write API calls and who also understand web services.

     

    AI is part of Atrium Core even though it is a separate installer...you do not need a separate license and it can handle all the data transformations you have listed and is very scalable.

  • 2. Re: Automated data population to CMDB
    Lin Zhu

    Hi dirk,

     

    Thanks for your reply! I will dig into the AtriumIntegrator7.6.04_UsersGuide right now to find out the best practice I could do.

     

    My another concern is that the current BMC CDM doesn't fit our requirements very well. Potentially we have to create new regular classes (many and maybe each with over 100 or even 200 attributes) to extends the current BMC CDM classes (mostly are BMC_Settings). Can Integrator overcome this ugly design to have a good performance as you mentioned "very scalable" (any potential risks)?

     

    I know the federation will be ideal way to tackle this ugly guy but we couldn't.

  • 3. Automated data population to CMDB
    Adrian Buckley

    Lin,

     

    Seamless Technologies has a product called the SeamlessData Pump that can handel your requirement ... you can find more information at www.seamlessti.com/products ... or contact me directly at sales@seamlessti.com. We are a BMC partner of 17 years and our products are certified by BMC.

     

    Regards

    Adrian

  • 4. Re: Automated data population to CMDB
    Lin Zhu

    Hi Abuckley,

     

    Thanks for your recommendation. The SeamlessData Pump sounds cool but our team doesn't have any plan to use any third-party solutions at this time. Thanks again!

  • 5. Automated data population to CMDB
    Laurent Matheo

    Another way that works is sending data directly from csv files using DataImport.bat.

    You can send data to (for example) form "bmc.core:bmc_computersystem" if you want to inject some Computer Systems. Also works for relations.