Fastdfs HOWTO-File and directory structure

The FASTDFS server-side runtime directory structure is as follows:
${base_path}
|__data: Storing Data files
|__logs: Storing log files

Where ${base_path} is set by the parameter "Base_path" in the configuration file.

One, tracker server

Tracker Server directory and file structure:
${base_path}
|__data
| |__storage_groups.dat: Storing group information
| |__storage_servers.dat: Storage Server List
|__logs
|__trackerd.log:tracker Server log files

The records in the data file Storage_groups.dat and Storage_servers.dat are separated by newline characters (/n), and the fields are separated by a Latin comma (,).
The fields in Storage_groups.dat are:
1. Group_name: Group name
2. Storage_port:storage Server port number

Storage_servers.dat Records Storage server related information in the following fields:
1. Group_name: Group name
2. Ip_addr:ip Address
3. Status: State
4. SYNC_SRC_IP_ADDR: To synchronize the source server of an existing data file to the storage server
5. Sync_until_timestamp: The deadline for synchronizing existing data files (Unix timestamp)
6. Stat.total_upload_count: number of files uploaded
7. Stat.success_upload_count: Number of successful uploaded files
8. Stat.total_set_meta_count: Change meta data times
9. Stat.success_set_meta_count: Successful change of meta data times
Stat.total_delete_count: Number of deleted files
Stat.success_delete_count: Number of files deleted successfully
Stat.total_download_count: Download number of files
Stat.success_download_count: Number of successful downloads of files
Stat.total_get_meta_count: Get meta data times
Stat.success_get_meta_count: The number of meta data successfully obtained
Stat.last_source_update: Last Source Update time (update operation from client)
Stat.last_sync_update: Last Sync update time (update operation from other Storage server synchronization)

II. Storage Server

Storage Server directory and file structure:
${base_path}
|__data
| |__.data_init_flag: Current Storage server initialization information
| |__storage_stat.dat: Current Storage server Statistics
| |__sync: Storing data synchronization related files
|     | |__binlog.index: Current binlog (Update action Log) file index number
|     | |__binlog.###: Store update action record (log)
|     | |__${ip_addr}_${port}.mark: To store the completion of synchronization to the target server
| |
| |__ Directory: 256 directories containing data files with hexadecimal characters such as: 1F
| |__ Level Two directory: 256 directory containing data files, the directory name is hexadecimal characters, such as: 0A, CF
|__logs
|__storaged.log:storage Server log files

The. data_init_flag file format is the INI configuration file, with each parameter as follows:
# Storage_join_time: This storage server creation time
# Sync_old_done: This storage server has completed the synchronization flag (the source server synchronizes existing data to this server)
# Sync_src_server: To synchronize the existing data to the server IP address of the source server, no is null
# Sync_until_timestamp: Sync existing data file up time (Unix timestamp)

The Storage_stat.dat file format is the INI configuration file, with each parameter as follows:
# Total_upload_count: number of files uploaded
# Success_upload_count: Number of successful uploaded files
# Total_set_meta_count: Change meta data times
# Success_set_meta_count: Successfully changed meta data times
# Total_delete_count: Number of deleted files
# Success_delete_count: Number of files deleted successfully
# Total_download_count: Download number of files
# Success_download_count: Number of successfully downloaded files
# Total_get_meta_count: Get meta data times
# Success_get_meta_count: Successful acquisition of Meta data times
# last_source_update: Last Source Update time (update operation from client)
# last_sync_update: Last Sync update time (update operation from other storage server)

There is only one data item in the Binlog.index: The file index number of the current Binlog

binlog.###,## #为索引号对应的3位十进制字符, less than three digits, the front of 0. The index number is based on 0 and the maximum is 999. A binlog file has a maximum of 1GB. Records are separated by newline characters (/n), and fields are separated by a Latin space. The fields are:
1. Timestamp: Update occurrence time (Unix timestamp)
2. Op_type: Operation type, one character
3. FileName: The file name of the operation (update), including the relative path, such as: 5a/3d/fe_93_sjz7paaao_bxyd. S

${IP_ADDR}_${PORT}.MARK:IP_ADDR is the target server IP address for synchronization, port is the storage server port for this group. For example: 10.0.0.1_23000.mark. The file format is the INI configuration file, with each parameter as follows:
# Binlog_index: The Binlog index number that has been processed (synchronized)
# Binlog_offset: Processed (synchronized) to Binlog file offset (in bytes)
# Need_sync_old: Sync Existing Data file tags, 0 means no data files need to sync
# Sync_old_done: Sync existing data file to complete tag, 0 indicates incomplete, 1 indicates completed
# Until_timestamp: Sync existing data up to time Point (Unix timestamp)
# Scan_row_count: Number of Binlog Records scanned
# Sync_row_count: Number of synchronized Binlog records

The data file name is automatically generated by the system, including three parts: the current time (the Unix timestamp), the file size (in bytes), and the random number. File name length is 16 bytes. Files are distributed in a PJW hash algorithm hash to 65536 (256*256) directories.