Putting new update-website script online ASAP

Jeremy Huntwork jhuntwork at linuxfromscratch.org
Sun Nov 14 07:58:18 PST 2004

Anderson Lizardo wrote:

>On Sunday 14 November 2004 07:49, Jeremy Utley (-J-), LFS Staff wrote:
>>Hey Guys!
>>Anderson forgot to update the update-website.sh script today when he
>>restructured the repository.  This caused the website regeneration script
>>to go haywire, and once again, almost the entire website got deleted.
>I'm really sorry for that. I thought I had changed all occurrences of the old 
>repos path before the reorganization but forgot the most important one (that 
>of the update-website.sh).
>>So, someone who understands the scripts better than I do needs to get the
>>website back into operational condition.  Then, we need to find out WHY
>>the website gets deleted every time one of these scripts malfunctions, and
>>resolve that.  I'd rather see the site not be generated at all, rather
>>than this happen!
>Basically, the website is deleted because the current running script does 
>something like
><commands to create a new website tree>
>cp new_website TARGETDIR
>So, if the website recreation fails somehow, the website is deleted because of 
>the "rm -rf" command. It's a badly designed procedure, which the new update 
>website script I'm working on plans to fix once and for all.
>Currently the new update-website.mk script (which only touchs 
>test.linuxfromscratch.org for now) is working well. The "pemission denied" 
>errors you saw were because you are not in lfswww group. I will change the 
>test website tree to "svnwww" onwership so everyone with commit privileges to 
>the www repos will trigger the update script properly.
>Because of this (unacceptable IMHO) new website massive deletion, I plan to 
>antecipate the new update script adoption for the current website. I'll just 
>fix the group onwership issue and then we can start the migration.
I think that would be great. We need to get that in place. Can you give 
me a bit more of a detailed walkthrough of the script?  I'd like to 
fully understand what it's doing so I can help if you aren't around for 
some reason.  It looks like it will backup the entire current site to 
/var/tmp before it generates the new one, is that right?  Just a 
thought, but could we (for the sake of redundancy) also consider pushing 
that backup (perhaps just that tar file) to one dedicated mirror for 
that purpose?  That way we are protected twice, in case the script fails 
or something major happens.  We could have it tar up the file, then 
attempt to send that file to a remote server.  If that fails, because 
the remote is down, or some other network problem, we just time-out and 
continue.  But if we are successful most of the time in our remote sync 
of this file, then we would likely always have the website rendered and 
stored somewhere that is at most a couple of days old.  Thoughts?

Jeremy Huntwork

More information about the website mailing list