Hi All,
I've set up some mirrors of madwifi.org which Mike suggested I post detai=
ls of=20
here.
There is a read-only mirror of the madwifi.org website at:
http://madwifi-mirror.locustworld.com
I've also set up a read-only svn mirror which can be accessed for example=
:
svn checkout svn://svn.locustworld.com/madwifi/trunk madwifi-ng=20
There is also an older snapshot mirror of tarball files here:
http://madwifi-files.locustworld.com/
The svn mirror syncs daily, the web mirror every few days and the snapsho=
t=20
files mirror will probably update weekly or monthly.
Best Regards,
Jon Anderson
On Tuesday 22 November 2005 16:57, Michael Renzmann wrote:
> Hi all.
>
> The recent downtimes sucked for everyone, I'm well aware of that fact.
> Although I'm confident that replacing the old disk solved the original
> problem I think it would be a good idea to have mirrors up and running
> which could act as temporary work-around in case the primary site is
> down and, as a nice side effect, could offer a better download speed in
> some cases.
>
> The most important question now is: which services should be mirrored,
> and in which way? So first we should have a look at the different
> services that madwifi.org currently offers:
>
> 1. Trac.
>
> This is an integrated environment for documentation (held in the Wiki),
> issue tracking (tickets), repository browsing and some simple project
> management capabilities (milestones). I think it's safe to say that thi=
s
> has become the heart of the project since it was started about one mont=
h
> ago. Trac is entirely database-driven, all content is stored in a local
> SQLite database. Trac itself is implemented in Python.
>
> 2. Subversion.
>
> The place for maintaining the code, including the complete history of
> everything that ever was in the driver. The repository is stored in
> Subversion's fsfs format. Subversion is implemented in Ansi C, and we'r=
e
> using it as module for Apache2.
>
> 3. Snapshots.
>
> Snapshots of the most recent code from trunk are usually generated once
> per day, if there has been changes to the repository since generation o=
f
> the last snapshot. Access to the archive is offered via Apache2's
> auto-generated directory indexes.
>
> Apart from these three services, we have an lxr running (automated
> web-based cross-reference for the latest code from trunk), as well as
> two IRC bots. I don't describe them here, since it makes little to no
> sense to mirror this type of stuff - they are not vitally important.
>
>
> The snapshots are quite easy to mirror. I could push them to every
> mirror as soon as a new snapshot has been generated, or the mirrors
> implement their own messures to sync themselves with the main site.
>
> Subversion is a bit trickier. First of all: svn mirrors should be
> read-only, else synchronization will be hell.
> There seem to be several tools and/or scripts that allow to keep
> multiple repositories synced. As far as I could see they seem to expect
> that the main site pushes the changes to the mirrors, but I could be
> wrong.
> Another option would be to rsync or download daily generated,
> incremental repository backups that then are loaded into the mirror
> repositories.
>
> Trac will be the hardest thing in terms of mirroring. As far as I found
> out so far, there are no tools for synchronizing mirrors. In addition,
> the mirrors should be read-only out of the same reason as for
> Subversion: synchronizing Trac changes among several sites will be hell=
=2E
>
> I'm not sure about how to mirror Trac. Maybe it would be best to
> generate a static (HTML) version of the Wiki content, and maybe also th=
e
> current open tickets. This would allow basically anyone to host a
> mirror, since then there is no dependancy to Python. In addition it
> would be a "natural" way to make sure that the mirror is read-only.
>
> Mirroring the full database probably is no good idea, since it contains
> sensitive information (passwords) which are of no use on the mirrors
> (since they won't provide write access to any Trac content). In
> addition, a full database mirror requires that the mirror makes use of
> Python - this unnecessarily rises the bar for potential mirror
> contributors.
>
>
> Comments?
>
> Bye, Mike
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by the JBoss Inc. Get Certified Today
> Register for a JBoss Training Course. Free Certification Exam
> for All Training Attendees Through End of 2005. For more info visit:
> http://ads.osdn.com/?ad_id=3D7628&alloc_id=3D16845&op=3Dclick
> _______________________________________________
> Madwifi-users mailing list
> Madwifi-users@...
> https://lists.sourceforge.net/lists/listinfo/madwifi-users
|