mydumper generates several files during the generation of the dump. Many of these are for the table data itself since every table has at least one file.
When a dump is executed a file called .metadata is created in the output directory. This contains the start and end time of the dump as well as the master binary log positions if applicable.
This is an example of the content of this file:
Started dump at: 2011-05-05 13:57:17
SHOW MASTER STATUS:
Log: linuxjedi-laptop-bin.000001
Pos: 106
Finished dump at: 2011-05-05 13:57:17
The data from every table is written into a separate file, also if the --rows option is used then each chunk of table will be in a separate file. The file names for this are in the format:
database.table.sql(.gz)
or if chunked:
database.table.chunk.sql(.gz)
Where ‘chunk’ is a number padded with up to 5 zeros.
When the --schemas option is used mydumper will create a file for the schema of every table it is writing data for. The files for this are in the following format:
database.table-schema.sql(.gz)
Binary logs are retrieved when --binlogs option has been set. This will store them in the the location specified in the --binlog-outdir option, by default this is the binlogs/ sub-directory inside the dump directory.
The binary log files have the same filename as the MySQL server that supplies them and will also have a .gz on the end if they are compressed.