How to tell if Tivoli is using utf8 support on Netware 6.5 No data sets can be found

No data sets can be found. This is the error that would appear in the Tivoli logs when backing up utf8 files with unmappable characters in them. They were using the 5.4.1 client which is the Unicode TSM client. Whether TSM is using UTF8 calls or not depends on if the customer has migrated previously existing file spaces that were backed up using MBCS calls to unicode file spaces. If the customer is still backing up to old file spaces, TSM will use the old MBCS codepage calls for filenames rather then unicode. If the files have unicode characters in them, this could cause translation problems if a unicode character does not exist in the codepage tables. This translation problem could result in an invalid filename, causing the tape software to skip the file as it does not exist in the file system under the incorrectly translated name. Hence the error: No data sets can be found.

If the customer creates a new file space using the unicode client, this will default to unicode file spaces causing the TSM client to use utf8 calls. TSM supports MBCS backwards compatibility because some customers are maintaining data for many years, in order to be compliant with tax laws for example. If there is a question on what TSM would be using, have the customer issue

query filespace f=d

on the TSM server console and check the file spaces that the TSM client is backing up to, to see if they report as Unicode Yes or No. Unicode yes is using UTF8. Unicode no is using the old MBCS calls.

-Make sure you have at least the 5.4.1 Tivoli client for Netware.

-Make sure you have the latest Novell libc updates.

-Use the query command above to see if Tivoli is using utf8 or MBCS calls. If the old MBCS calls are being used, you may have to create a new filespace using the 5.4.1 or later Tivoli client. You may need to contact IBM for further information on creating the new filespace.