<html>
<head>
<meta content="text/html; charset=windows-1252"
http-equiv="Content-Type">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<br>
<br>
<div class="moz-cite-prefix">On 04/01/2015 05:22 AM, Simon Barrett
wrote:<br>
</div>
<blockquote
cite="mid:CAMOhFjHhxWuw8zez9cNKW1+-_u7D=mg8zBmMdr4Ord627h8x5w@mail.gmail.com"
type="cite">
<meta http-equiv="Content-Type" content="text/html;
charset=windows-1252">
<div dir="ltr">Hi all,
<div><br>
</div>
<div>Is there any reason why I should not create snapshot a
ploop-backed container each day (or hour, for that matter)
then merge all outstanding snapshots at the end of the week
(vzctl snapshot-delete) and compact it? This would allow me
to do efficient incremental backups to back up (using bacula,
in this case) and I would imagine it's a core use case.</div>
<div><br>
</div>
<div>Are there any potential risks to be aware of (performance
issues, storage usage, corruption exposure)?<br>
</div>
</div>
</blockquote>
<br>
I am not aware of any, except a limitation of no more than 126
deltas mounted<br>
(i.e. the distance from top to base delta should be < 127). If
you try to hit the limit,<br>
snapshot won't be created. So, hourly snapshots with weekly delete
won't work, as<br>
24*7 > 127.<br>
</body>
</html>