3 ms·
Except that your one EC2 instance is a single point of failure, from instance through host through availability zone, and that providing reliable, multi-node NF
by jdub 12y ago
Except that your one EC2 instance is a single point of failure, from instance through host through availability zone, and that providing reliable, multi-node NFS is hard (even with GlusterFS).
And you have to resize your block devices to store more stuff.
And...
And...
:-)
- nkozyra 12y agoWell, this is a problem with a lot of ephemeral services hand-spun on AWS. There are also ways of mitigating it on your instances, but those are clunky, too. Which is why new services like these are always preferable.