hi all

as it says in the title
when i archive a file, should i make a copy of the file, then tar and gzip the copy
or can i just tar and gzip the original file

what i am trying to do is backup my web server, eg. httpdocs
eg. tar -cf httpdocs_date.tar /httpdocs
gzip -9 httpdocs_date.tar

or

cp /httpdocs /httpdocs_date
tar -cf httpdocs_date.tar /httpdocs_date
gzip -9 httpdocs_date.tar

with speed, impact on web site, restoration, and data integrity as the key points

would it just be best to shutdown the server while this is being done?

the directory i'm guessing will be massive, several gigs at least

thanks
dave

If httpdocs is static (not files written to it by client-sessions)
the tar-ing of the actual thing should be just fine. If you
have the content modified even a copy wouldn't prevent
broken files necessarily - you'd have to stop the service(s),
make your copy and restart them.


Cheers,
Tink

hey thanks for your reply
so if i stopped the service would i still need to make a copy to tar?

Nope - the only "problem" is that a tar with bz2
compression would make for a longer downtime
than a copy. If you tar, restart the service, and
THEN compress you should be sweet :}


Cheers,
Tink