Search outside of database: facing setup issues

I need to use "Search outside of Database" implementation for my requirement. I have table with id, file_path, created_on columns.

I though of implemnting following DumbFetcher

public class DumbFetcher extends AbstractFetcher {

    @Override
    public void execute(Properties p, long lastRunTime, List<Column> columns) {

1. . Execute: select id, file_path, created_on from files_list
2. . For each record
3. . <<FileTypeFilter>> ft = new <<FileTypeFilter>>()
4. . ft.add("id", id);
5. . ft.add("file_stream", new <<Custom_API_to_get_String>>(file_path).toString());
6. . ft.add("created_on", created_on);
7. . scheduleDocument(ft);

	}	
    @Override
    public List<FieldType> getFieldTypes(Properties p) {
        List<FieldType> ret = new ArrayList<FieldType>();
        ret.add(new NumberFieldType("id").setPrimaryKey(true));
        ret.add(new StringFieldType("file_content"));
        ret.add(new TimestampFieldType("created_on").setModifiedTime(true));
        return ret;
    }	

1. Is the Above implementation correct?

2. I couldnt understand the use of List<Column> columns argument to execute method. Please explain or guide me to reference document.

3. "Code line 3" should use <<FileTypeFilter>> (for example DocTypeFilter, CvsTypeFilter etc..) or always should use TextDocument? How to integrate new File Type Filters to the DumbFetcher?

4. "Code line 5" should always pass content of the file or does it accept FileTypeFilter?

5. How to handle transaction i.e. if any document failed indexing how to reindex it? is it taken care by DBSight/lucene or my DubmbFetcher should have implemented with handler for fail over mechanism?

6. How to inform lucene about a document delete from the source?

7. While configuring Index through admin UI facing few problems. Even after selecting CustomFetcher, index configuration screens keep asking main query and sub query. i feel "Search out side DB" shouldn't ask this configuration. Do you have any screen casts on how to setup search outside DB?

8. Why primary key should be of keyword type? I don't want my primary key to be searched since its a internal unique number and user shouldn't know about it. As per description keyword is also included in search index

9. Last step giving following message "Scheduling indexing is not included in Free License Level. Please upgrade to allow DBSight to work for you". How do i configure custom fetcher and go ahead with indexing and search?

Thanks for your time.