Friday, December 21, 2012

Gaining more grip over repo sync problems

Hi this document is targeted to develop a C code which provides much efficiency in downloading huge files from repositories. The authors of the document is Tirtha Ghosh and A G Megharaj. If interested to access the information please do contact the authors.

The basic idea behind the development of this software is that, sometimes we may face problems in downloading especially with large files when there is power off. So to over come this the software actually takes the name of the valid projects and saves in a file and keep track of the download. Giving more grip over repo sync.

To be simple the software works as explained below

The C-code will run "repo sync " rather "repo sync". When Internet speed is an issue we have faced so many times repo hung and it started back all from begining.

This C code will read the manifest.xml file and create the project name list and then do repo sync on each project name and once downloaded marked them as DONE. Once a project name is marked DONE, "repo sync " will not be performed for those.

Usage:

./repo-sync all/create/sync filename

Note:
1) filename is optional for all but mandatory for create and sync. In this file project names will be stored.
2) create is for creating filename with project names.
3) sync is for syncing projects one by one and mark downloaded one as DONE in filename.
4) all is for doing both back to back (filename is optional, default name is projects.txt)

you can download the binary from the link below.

https://docs.google.com/open?id=0B1seuVAfI1ktU19UOHVHUzRlSWM

No comments:

Post a Comment