Use correct fSeq after disk read error

When the tape write session encounters a corrupted disk file, it should skip it and continue with archiving the remaining files.

As seen in https://gitlab.cern.ch/cta/operations/-/issues/934, next fSeq is calculated incorrectly. A developed should look inside session->validateNextFSeq(m_archiveJob->tapeFile.fSeq), find where fSeq is incremented and ensure it is correct in case of disk read errors.