Automatic website backup

Anderson Lizardo lizardo at linuxfromscratch.org
Sun Feb 13 16:05:09 PST 2005


On Sunday 13 February 2005 19:33, Matthew Burgess wrote:
> That's why I mentioned timestamping the tarball names.  That way you
> keep, say, 10 tarballs around, and only the last one or two would have
> an empty tree.  The main issue I have though is why someone feels it
> necessary to have a script that removes stuff from the website anyway,
> but then I never did look at the website generation scripts.

That's how the current running script (update-website.sh) works:

1) render the entire website on temp_dir
2) "rm -rf /home/httpd/www.linuxfromscratch.org"
3) move the new website content from temp_dir 
to /home/httpd/www.linuxfromscratch.org
4) rm -rf temp_dir

The script running on test.linuxfromscratch.org (update-website.mk) works by 
keeping an "opaque" SVN working copy of www+patches+hints, rendering the 
website inside this working copy (not the entire website, just what actually 
needs to be re-rendered) and the rsync'ing 
to /home/httpd/www.linuxfromscratch.org. So there is no "rm -rf" envolved.

If I understood correctly, you have beta.linuxfromscratch.org working like 
test.linuxfromscratch.org, right?
-- 
Anderson Lizardo
lizardo at linuxfromscratch.org
http://www.linuxfromscratch.org/




More information about the website mailing list