On Nov 1, 2011, at 05:24, Nrupen Kantamneni wrote:

> 1.       Can we have revision control via web browser based ? If yes can you 
> provide me the settings

What functions do you want to make available via the web?

Out of the box, if you serve your repository with Apache, you get a web-based 
view of the files and directories in your repository, including past revisions 
if you know what URL parameters to append. You can customize the appearance of 
directory listings using an XSLT stylesheet.

If you install third-party web-based Subversion repository viewers like Trac, 
WebSVN or ViewVC, you get a prettier view, including additional functions like 
examining log messages and diffs, and then it doesn't matters how your 
repository is served (i.e. it no longer needs to be served by Apache; it could 
be served by svnserve).

There are also third-party web-based Subversion administrative programs like 
SVNManager that let you create users and repositories.

I'm not aware of any web-based interfaces for modifying the contents of a 
repository, such as creating, editing, renaming, moving or deleting files or 
directories. Typically you want to make changes in a working copy so that you 
can test your changes before committing them.


> 2.       Does SVN support hot swap replication (meaning the data is 
> continuously backed up on a different machine and if the original server goes 
> down the secondary immediately start serving as master)

You can create replica servers using svnsync. These can be used purely as 
offline backups, or can be made available online and used simultaneously, as 
long as they are configured to be read-only. You may only commit changes to the 
master repository, but you can configure the slave repositories to 
automatically proxy write requests back to the master. If the master fails, 
manual intervention is necessary to promote one of the slaves to be the new 
master. Or if the failure of the master is temporary and having only read 
access temporarily is acceptable, you can just bring the master back up when 
you can, and let users read from one of the slaves until then. Users do however 
need to specifically select which server they want to connect to (whether the 
master or a specific slave), so regardless of their choice, if that particular 
server is down, those users will have to "svn relocate" their working copies to 
another server if they need access to the repository during the time that 
server is down.




Reply via email to