CP855 is a Cyrilliccodepage to be used under MS-DOS. This codepage is not much used. The Cyrillic alphabet (or azbuka, from the old name of the first letters) is an alphabet used to write six natural Slavic languages (Belarusian, Bulgarian, Macedonian, Russian, Serbian, and Ukrainian) and many other languages of the former Soviet Union, Asia and Eastern Europe. ... Code page is the traditional IBM term used for a specific character encoding table: a mapping in which a sequence of bits, usually a single octet representing integer values 0 through 255, is associated with a specific character. ... Microsofts disk operating system, MS-DOS, was Microsofts implementation of DOS, which was the first popular operating system for the IBM PC, and until recently, was widely used on the PC compatible platform. ...
In the following table only the codes 128–255 are shown. The codes 0–127 are the same as in ASCII. There are 95 printable ASCII characters, numbered 32 to 126. ...
Java uses a default encoding which changed in versions later than 1.1.7 from iso-8859-1 to CP1252 (I assume that for Cyrillic that translates to iso-8859-5 and cp1251, respectively).
Windows and DOS have a whole list of changes with respect to Cyrillian encodings: CP855, CP866 and finally CP1251, which is incompatible with both iso-8859-5 and ISO-IR-111 (aka KOI8).
Since your characters are displayed as '?' it's very unlikely you'll see a UTF-8 encoding in your output file (with all due respect), it rather seems to be a mismatch between one of the iso encodings and CP1251, that's my guess.