[Openstack] Swift large data ingress thoughts?

Jon Slenk jslenk at internap.com
Thu May 12 21:03:51 UTC 2011


hi,

Any experience or recommendations or thoughts about getting big data
into Swift in the first place? I am figuring there are options like:

a) just do it over the public network: slowest but most easy for everybody.
b) go sneakernet, and ship a physical device like a single fat disk,
or raid, or lots of ssds, and hook those up on the internal Swift
network and send them through the proxy: faster but requires some sort
of real human service org inside Swift.
c) [crazy idea, i have no idea how to get this to work, even if it
isn't a dumb idea] directly attach the disk as an object store so that
the files are immediately serviceable without any replication in
place, and let the automatic gradual replication spread it out to the
full cluster?

thanks for any insights.
sincerely,
-Jon.




More information about the Openstack mailing list