do-snapshot.sh

#!/bin/sh

#
# create a copy of the ESGF mirror site, consisting of hard links
#
# make it in a temporary location, and if it succeeds, rename it 
# to a dated directory under the webserver (outside of the main tree)
#
# otherwise, leaves it in tmp location to be cleaned up next time
# this cron job runs.


top=/path/to/htdocs

maindir=$top/esgf
snapdir=$top/esgf_snapshots
tmpcopy=$snapdir/.partial_copy

rm -fr $tmpcopy

cp -Ral $maindir $tmpcopy
status=$?

if [ $status -eq 0 ]
then
    dest=$snapdir/`date +%Y%m%d`
    if [ -e $dest ] ; then dest=$snapdir/`date +%Y%m%d_%H%M%S`; fi
    
    mv $tmpcopy $dest
fi

Script to make snapshots of esgf distribution on the rsync mirror sites.

Set the path in line 13, run daily from cron, and use script at https://gist.github.com/alaniwi/e9186a8b9897c77b1fcc8adf8d25b6dc to delete some older snapshots to reduce this to weekly.