Export Network Device issue

[alyautdinov]

Hello Team

We want to migrate from test deployment to production deployment and create separate ISE instance.

We export "network device" from test, but we can't import them to production deployment.

The issue is the parameters is exported with the wrong characters

The example in the attachment

The problem is existing with devices which was migrated from ACS or duplicated.

How can we fix this behavior?

ISE version 2.3 patch 2

[Craig Hyps]

Are you saying the template has wrong characters, or values for devices have unexpected characters?

If the template is incorrect, then could adjust headers, but if exporting to/from same version, typically would not need to change.  May need to open TAC case to resolve, especially if issues with migration tool.

[alyautdinov]

Values for devices have unexpected characters.

In example instead numerical value (2083) we get symbols "NaN"

[Craig Hyps]

Does it appear in Admin UI with value of 2083 or other characters?  If appears in ISE as 2083 and exports as different value, then need TAC case to troubleshoot.  If value appears as foreign characters, then issue with migration tool.

[hslai]

By reading the CSV file, it appears the export came from ISE 2.4 beta build.

If you like us to look further into this issue, please provide a sample NAD CSV from ACS, ACS release and patch number, and ISE shipping release and patch number used.

I see this column "DtlsCoaPort" is wrong in your CSV:

DtlsCoaPort:Integer(128)

NaN

[alyautdinov]

ACS version:

Version : 5.8.0.32.7
Internal Build ID : B.442
Patches :  5-8-0-32-4
            5-8-0-32

ISE version: 2.3 patch 2

[hslai · Accepted solution]

Was the ACS instance migrated from some older ACS release, because it has a NDG hierarchy of Migrated_NDGs?

I tried importing your CSV file to ACS 5.8 and migrating the NDGs and the NADs from ACS 5.8 to ISE 2.3. After that, NAD export did not get NaN but an empty column.

If you are after a quick fix, you may open the exported CSV in MS Excel or similar CSV editor, replace any incorrect values and re-import the modified CSV to your test deployment. Once the import succeeds and some sanity tests go OK, then the file can be used in production.