ndncatchunks and ndnputchunks

ndncatchunks and ndnputchunks are a pair of programs to transfer a file as Data segments.

• ndnputchunks is a producer program that reads a file from the standard input, and makes it available as NDN Data segments. It appends version and segment number components to the specified name, according to the NDN naming conventions.

• ndncatchunks is a consumer program that fetches Data segments of a file, optionally discovering the latest version of the file, and writes the content of the retrieved file to the standard output.

Version discovery methods in ndncatchunks

• fixed : sends an interest attempting to find a data packet with the specified prefix and version number. A version component must be present at the end of the user-specified NDN name.

• iterative: sends a series of interests with ChildSelector set to prefer the rightmost child and Exclude selectors, attempting to find a data packet with the specified prefix and the latest (the largest in the NDN canonical ordering) version number. The version is declared "latest" after a predefined number of data retrieval timeouts (default: 1).

The default discovery method is iterative.

Interest pipeline types in ndncatchunks

• fixed: maintains a fixed-size window of Interests in flight; the window size is configurable via a command line option and defaults to 1.

• aimd : sends Interests using an additive-increase/multiplicative-decrease (AIMD) algorithm to control the window size. By default, a Conservative Loss Adaptation algorithm is adopted combining with the AIMD algorithm, that is, at most one window decrease will be performed per round-trip-time. For details please refer to: A Practical Congestion Control Scheme for Named Data Networking

• cubic : sends Interests using the TCP-Cubic algorithm to control the window size. For details of the TCP-Cubic, please refer to: CUBIC: A New TCP-Friendly High-Speed TCP Variant

The default Interest pipeline type is fixed.

Usage examples

Publishing

The following command will publish the text of the GPL-3 license under the /localhost/demo/gpl3 prefix:

ndnputchunks ndn:/localhost/demo/gpl3 < /usr/share/common-licenses/GPL-3

To find the published version you have to start ndnputchunks with the -p command line option, for example:

ndnputchunks -p ndn:/localhost/demo/gpl3 < /usr/share/common-licenses/GPL-3

This command will print the published version to the standard output.

To publish data with a specific version, you must append a version component to the end of the prefix. The version component must follow the aforementioned NDN naming conventions. For example, the following command will publish the version %FD%00%00%01Qc%CF%17v of the /localhost/demo/gpl3 prefix:

ndnputchunks ndn:/localhost/demo/gpl3/%FD%00%00%01Qc%CF%17v < /usr/share/common-licenses/GPL-3

If the version component is not valid, a new well-formed version will be generated and appended to the supplied NDN name.

Retrieval

To retrieve the latest version of a published file, the following command can be used:

ndncatchunks -d iterative ndn:/localhost/demo/gpl3

This command will use the iterative method to discover the latest version of the file.

To fetch a specific version of a published file, you can use the fixed version discovery method. In this case the version needs to be supplied as part of the name. For example, if the version is known to be %FD%00%00%01Qc%CF%17v, the following command will fetch that exact version of the file:

ndncatchunks -d fixed ndn:/localhost/demo/gpl3/%FD%00%00%01Qc%CF%17v

To retrieve the file with aimd pipeline:

ndncatchunks -t aimd -d iterative ndn:/localhost/demo/gpl3

To retrieve the file with cubic pipeline:

ndncatchunks -t cubic -d iterative ndn:/localhost/demo/gpl3

For more information, run the programs with --help as argument.