Hello!
I'm helping my organization create a CMDB and Asset Management practice. When I arrived, there were approximately 500 items in CMDB that were essentially a record name with no metadata, originally added to support other integration efforts that needed to reference a device. Now that we've built in some discovery capabilities, most of these items have been reidentified and reclassified to the correct tables.
Now I'm planning to load a manual list of legacy network switches to clean up the rest of the original data. However, here is our situation / issue:
- We only have current CMDB records for these devices with a name and no additional metadata
- In our source data sheet, we have multiple devices with the same name and IP as a "stack" with different serial numbers
- Using the base OOB identification rules, when we first loaded this sheet using ETL, we only wrote to 98 of the designated 130 records. Many of the records with duplicate host names would not write a record. Essentially, the ETL would find the first record with the same name, give a serial, and then skip any records from the source sheet with the same name
- So, to try and fix this we created individual records for each device with duplicate names and added the serial # for those duplicate names. We should now have an individual record 1:1 to every item in our source list. However, again, only 98 of 130 records....
We aren't sure what to do here. Anyone have recommendations on how to handle the identification rules to allow multiple records with the same name while having different serial numbers and to ensure that multiple records would be created from the source data? We are new to this, so any help and guidance would be a huge support!
