Preprocessor Documentation


File services

Hui Cao

########################################## # THIS CODE IS STILL EXPERIMENTAL! # DO NOT USE IN PRODUCTION ENVIRONMENTS. # Please send any issues to the Snort team ##########################################


With the volume of malwares transferred through network increasing, network file inspection becomes more and more important. This feature will provide file type identification, file signature creation, and file capture capabilities to help user deal with those challenges.

There are two parts of file services: file APIs and file inspect preprocessor. File APIs provides all the file inspection functionalities, such as file type identification, file signature calculation, and file capture. File inspect preprocessor provides users ability to control file services, such as enable/disable/configure file type identification, file signature, or file capture.

Supported protocols: HTTP, SMTP, IMAP, POP3, FTP, and SMB. Supported file signature calculation: SHA256

Sections: Dependency Requirements File Services Configuration File Inspect Preprocessor Configuration File Type Identification File Signature File Capture

Dependency Requirements

For proper functioning of the file processing:

Stream session tracking must be enabled, i.e. stream5. TCP must be
enabled in stream5. The preprocessor requires a session tracker to keep 
its data. 

IP defragmentation should be enabled, i.e. the frag3 preprocessor should be
enabled and configured.

At lease one of preprocessors supported protocols must be enabled.

File Services Configuration

File services are configured through snort “conf” keyword. The configuration name is “file”:

config file:

Option Argument Required Default file_type_depth No 1460 bytes file_signature_depth No 10485760 bytes file_block_timeout No 1 day file_capture_memcap No 100 megabytes file_capture_max No 1048576 bytes file_capture_min No 0 bytes file_capture_block_size No 32768 bytes

depth = 0 - 4G bytes timeout = 0 - 4G seconds memcap = 1 - 4G megabytes max = 0 - 4G bytes min = 0 - 4G bytes size = 8 - 4G bytes

config file:\ < file_type_depth depth >,\ < file_signature_depth depth >, \ < file_block_timeout timeout >, \ < file_capture_memcap memcap >, \ < file_capture_max max >, \ < file_capture_min min >, \ < file_capture_block_size size >


< file_type_depth depth >: Maximum file depth to identify file type. If file type is not identified beyond this value, file type will be set “unknown”. 0 means unlimited

< file_signature_depth depth >: Maximum file depth to calculate file signature. If file size is larger than this value, file signature will not be calculated. 0 means unlimited

< file_block_timeout timeout >: When a file transferred through HTTP is blocked , browser might resume file transfer with partial content in HTTP protocol. Snort block file even if it is transferred through resume. This timeout is to set how long it will keep blocking that file.

< file_capture_memcap memcap >: This sets the memory limit for file buffers, in megabytes.

< file_capture_max max >: Maximum file size we can capture. If file size is greater than this value file will not be captured.

< file_capture_min min >: Minimum file size we can capture. If file size is smaller than this value file will not be captured.

< file_capture_block_size size >: This sets the memory block size for file buffer. Smaller value might save memory but may hurt performance. Default is 32k bytes.

Example: config file: file_type_depth 16384, file_signature_depth 10485760, \ file_block_timeout 3600, file_capture_memcap 200, \ file_capture_max 1048576, file_capture_min 200, \ file_capture_block_size 65536,

This set the file type depth 16K bytes, file signature depth 10M bytes, file block timeout 1 hour, file capture memcap 200M bytes, maximum file size to capture is 1M file, minimum file size to capture is 200 bytes, file block size is 64k.

File Inspect Configuration

File inspect dynamic preprocessor can be enabled during build time. The following options must be included in ./configure:


The configuration name is “file_inspect”:

Option Argument Required Default type_id NONE No DISABLED signature NONE No DISABLED capture_memory NONE No DISABLED capture_disk

No DISABLED capture_network No DISABLED capture_queue_size No 3000 blacklist No NULL greylist No NULL

portno = 0 - 65535 size = 0 - 4G files disk size = 0 - 4G megabytes

preprocessor file_inspect:\ < type_id >,\ < signature >, \ < capture_memory >, \ < capture_disk dir >, \ < capture_network hostname portno >, \ < capture_queue_size size >, \ < blacklist file>, \ < greylist file> Options:

< type_id >: Enable file type identification

< signature >: Enable file signature calculation

< capture_memory >: Capture file to memory (file buffers). File analysis can be done on file buffers directly without the overhead of storing to disk.

< capture_disk dir >: Store files to the directory specified in dir, and capture no more than disk size (in megabytes). If this limit is reached, no more file capture.

< capture_network hostname portno >: Send files to other host, specified by hostname (or IP address) and port number.

< capture_queue_size size >: Set the maximum number of files can be queued to process (saved to disk or sent to network).

< blacklist file >: Specify a SHA256 list file. Files match SHA256 will be blocked.

< greylist file >: Specify a SHA256 list file. Files match SHA256 will be alerted.

SHA list file format one SHA per line. Use # for comments.

#start of block.txt
#end of block.txt

Example: preprocessor file_inspect: type_id, signature, \ capture_disk /home/file_capture/tmp/ 300, \ capture_network localhost 8000, \ capture_queue_size 5000, \ blacklist block.txt

File Type Identification

File magic rule:

file type: ; id: ; ver: ; category: ; \ group: [,]; msg: ; \ content: ||; offset:; rev:

type: defines the name of file type. Name should be limited to any alphanumeric string including periods, dashes, and underscores.

id: used to uniquely identify different file types, similarly to rule ID.

ver: used to indicate different versions for the same file type.

category: defines the categories of file type. Name should be limited to any alphanumeric string including periods, dashes, and underscores. Categories can be Executables, PDF files, FLASH files, Office Documents, Archive, Graphics, Multimedia etc.

group: logical grouping of file rules, used for IPS rule keyword file_group. See README.file_ips for more information on ‘file_group’.

msg: the message to print when file type information needs to be described.

content: defines the file magic using binary data. The binary data is enclosed within the pipe (|) character and represented as bytecode, i.e hexadecimal numbers.

offset: defines the location of file magic relative to the start of file.

rev: used to uniquely identify revisions of file type definition.

Similar to snort rule, it can be used through include directives. Similar file types can be put into the same file. Multiple file type configurations can be included in snort configuration.

Pre-packaged file magic rules:

A set of file magic rules is packaged with Snort. They can be located at “etc/file_magic.conf”. To use this feature, it is recommended that the these pre-packaged rules are used; doing so requires that you include the file in your Snort configuration as such:

include etc/filemagic.conf


file type:GIF; ver:87; category:graphics; msg:”GIF87a”; id: 1; content: |47 49 46 38 37 61|; offset:0; rev:1; file type:GIF; ver:89; category:graphics; msg:”GIF89a”; id: 1; content: |47 49 46 38 39 61|; offset:0; rev:1;

The previous two rules define GIF format, because file magics are different.

Full set of file type magic rules are not included. Future release will include file type magic and rule options to use file type information.

Enable file type

File type events are generated when file type preprocessor rules are enabled. Each file magic rule requires one file type preprocessor rule. File type use 146 as generator ID.


alert ( msg: “GIF87a file “; sid: 87; gid: 146; rev: 1; metadata: rule-type preproc; ) alert ( msg: “GIF89a file “; sid: 89; gid: 146; rev: 1; metadata: rule-type preproc; )

preprocessor file_inspect: type_id

File Signature Calculation

File signature events are generated when file signature preprocessor rules are enabled. File signature use 147 as generator ID.


alert ( msg: “File signature “; sid: 1; gid: 147; rev: 1; metadata: rule-type preproc; )

Enable Openssl support

File services provide its own SHA calculations, but users can optionally enable openssl implementation. Openssl might provide performance optimization based on the platform snort runs. To enable it, the following options should be included in the ./configure command.

–with-openssl-includes= --with-openssl-libraries=

Example: –with-openssl-includes=/usr/local/ssl/include –with-openssl-libraries=/usr/local/ssl/lib

File Capture

File capture relies on file type and file signature enabled. If file type is unknown, that file will not be captured. File signature is used as file ID to avoid downloading duplicated files.

Example: preprocessor file_inspect: type_id, signature, \ capture_disk /home/file_capture/tmp/, \ capture_queue_size 5000

All files captured will be saved to directory “tmp”. File capture queue is set to 5000 files.

Sending captured files to another host

File inspect preprocessor provides a simple way to sending files to another host for further analysis. In tools/ directory, file_server is provided to receive and save all the files to disk.

Example: preprocessor file_inspect: type_id, signature, \ capture_network 8000, \ capture_queue_size 5000

All files captured will be sent to host, to port number 8000

On the host, you can start file_server before starting snort:

 file_server 8000

All the files will be saved to the directory where the command is running.

File Events

File inspect preprocessor also works as a dynamic output plugins for file events. It logs basic information about file.


output filelog: [ []]

::= [(’G’|’M’|K’)] filename: the name of the log file. The default name is /file. You may specify ”stdout” for terminal output. The name may include an absolute or relative path. limit: an optional limit on file size which defaults to 128 MB. The minimum is 1 KB. Example: dynamicoutput file / path to / output filelog:file All file events will be logged in "file" in the log directory. File event example: 07/02-16:42:50.253157 [**] [File: /wp-content/themes/images/logo.jpg, size: 4633 bytes] [signature: 8D68FA32E67624D409247FEF10D32A413D55C125970AFA533B9CA5E7DDF07A5E] [**] [147:1:1] [**] {TCP} ->