On Tue, Apr 8, 2008 at 3:14 PM, Chris McAvoy <<a href="mailto:chris.mcavoy@gmail.com">chris.mcavoy@gmail.com</a>> wrote:<br><div class="gmail_quote"><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<div class="Ih2E3d">
Sort of, it's just not persistent. If you shut down your instance,<br></div>
you lose your data. Unless there's something new I haven't heard of.<br>
<div><div></div><div class="Wj3C7c"><br>
Chris</div></div></blockquote><div>One of the big guys on Slashdot addressed comments of a similar type about EC2 here:<br>
<br>
<a href="http://developers.slashdot.org/comments.pl?sid=501678&cid=22883434">http://developers.slashdot.org/comments.pl?sid=501678&cid=22883434</a><br>
<br>
Basically, there's nothing stopping you from running a SQL server on
your instances, but if you're in the enterprise space with many
customers you shouldn't be hoping your single point of failure won't
fail. In this case, just set up a cluster of databases and send your
backups to Amazon S3 and that's how you'd handle it in a regular data center if you owned the physical servers. <br>
<br>
It's not trivial (high availability never is, *sigh*), but if you are running a single
SQL server instance and your hardware fails your data's gone anyway,
whether or not your hardware is virtual or in meatspace, so hopefully a
solid backup solution is in place.<br>
<br>
-Tim Gebhardt<br>
<a href="mailto:tim@gebhardtcomputing.com">tim@gebhardtcomputing.com</a> <br></div></div>