Backing up websites (large number of small files)
79 views
Skip to first unread message
Dave Warren
Jan 31, 2017, 5:41:48 PM1/31/17
to zba...@googlegroups.com
Howdy!
I'm exploring using ZBackup for a small web server and I'm trying to wrap my head around how the deduplication and bundles and indexes work.
We have a dozen sites that run WordPress, normally the current version although they may drift in versions slightly over time. A lot of other assets (ranging from KB sized PHP, JS, CSS files to 400MB media files) are shared between sites but are loaded to each site individually, so deduplication should be beneficial. Total storage on the server across all sites is only around 8GB right now (60000 files), we expect this to approximately double over the next year.
However, noting that restoring cannot restore an individual file and instead only restores an entire uncompressed tar, I wonder what is the best way to handle the typical use case where we need to retrieve just one site (or more typically, just a file within a site)?
Could I sequentially backup different sites like this:
tar c /var/www/html/$SITENAME | zbackup /zbackup/www/backups/$SITENAME-$DATE
In the future if I need to restore a file, I would still need to restore an entire site, but at least not the entire server.
Would I still get the deduplication benefits across sites? Is there a better way to accomplish this?
I'm exploring using ZBackup for a small web server and I'm trying to wrap my head around how the deduplication and bundles and indexes work.
We have a dozen sites that run WordPress, normally the current version although they may drift in versions slightly over time. A lot of other assets (ranging from KB sized PHP, JS, CSS files to 400MB media files) are shared between sites but are loaded to each site individually, so deduplication should be beneficial. Total storage on the server across all sites is only around 8GB right now (60000 files), we expect this to approximately double over the next year.
However, noting that restoring cannot restore an individual file and instead only restores an entire uncompressed tar, I wonder what is the best way to handle the typical use case where we need to retrieve just one site (or more typically, just a file within a site)?
Could I sequentially backup different sites like this:
tar c /var/www/html/$SITENAME | zbackup /zbackup/www/backups/$SITENAME-$DATE
In the future if I need to restore a file, I would still need to restore an entire site, but at least not the entire server.
Would I still get the deduplication benefits across sites? Is there a better way to accomplish this?
Reply all
Reply to author
Forward
0 new messages