Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.
Originally posted by: aorlovskyAndrey,
The message in the error log will go away if you set the property ls.brain.node.characterSet = "windows-1251"
The filter in you sample graph will still fail. This is because some of your column names are unicode characters. It will work just fine in BRAIN if you use the "windows-1251" encoding for your column names. That is, column names must not be written to the brd using utf-8, but rather as Cyrillic characters in the windows-1251 codepage. If you make this change, you will also have to specify windows-1251 as the metadata codepage for the BrdViewer.
The rules for using Unicode in BRAIN are available in the BRAIN Release Notes.
It would be helpful if you could also attach a small sample db so we could reproduce your import step exactly.
Were you using BRAIN 4.1.4?
Best regards