[GSoC 2017] New Idea about rsync

[Shanker Wang]

Hi, there

Hope this e-mail finds you well. I'm interested in GSoC program. And
I have an idea about the tool rsync.

I've been one of the admins of TUNA mirror site (mirrors.tuna.tsinghua.
edu.cn), which provides mirrors for a large amount of open source 
softwares. Late last year, as one of the top two mirror sites in China,
rsync request from other sites made the load on our storage system 
grow largely. I realized that to initialize a rsync connection, the 
requested directory would be scanned recursively, which was making 
massive IO operations and causing cache miss. To reduce the stress, 
I did some modification to `rsync` and was able to cache the attributes 
of the files. The result was fine, and is published on 
https://github.com/tuna/rsync and further documentation is on  
https://github.com/tuna/rsync/blob/master/README-huai.md

The modification itself was successful. During the process of the modifi-
cation, however, I found rsync is lack of documentation, especially the
communication protocol. [https://rsync.samba.org/how-rsync-works.html]
was one of a few documents for my reference. I noticed that samba is
participating the 2017 GSoC project and rsync is maintained by you. As
a result, I wonder if I can do something for the documentation for the
rsync protocol.

The achievement of this idea, as I expect, will include a detailed document
about the whole rsync protocol, which can be relied on if someone is 
working on another implementation of the rsync. I will document the 
latest protocol, as well as historical versions if there is time left.

I wonder if this idea is appropriate for a summer project. I also need 
your suggestion on my idea.

Cheers,

Miao Wang
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 3589 bytes
Desc: not available
URL: <http://lists.samba.org/pipermail/samba-technical/attachments/20170315/4f6e18c0/smime.bin>