On Sat, 28 Jan 2012, Matt Domsch wrote:
Responding here, but assuming that future messages will probably move to
the cloud@ mailing list.
I'm here. I think it's a combination of people in Cloud SIG with S3
experience, and Infra that's on the hook to deliver.
I put some thoughts as to how it could/should work here:
http://fedoraproject.org/wiki/Infrastructure/Mirroring/Amazon
but there are a lot of details about working with S3 that I simply
don't know (e.g. the per-bucket policies, and how to get the list of
IP netblocks for each region).
That is all stuff that I, or someone in my team, can help with, since
part of our mandate is to help external Linux distros to use AWS in
whatever manner they desire.
Regarding the IP addresses -- one of the things that we do with the
repositories for the Amazon Linux AMI (which are in every region's S3)
is place ACLs on those S3 buckets such that they will only accept
connections from the IP address ranges that are known to be EC2
instances. In other words, we ensure that only EC2 instances can access
the repositories.
Those IP ranges are here --
https://forums.aws.amazon.com/ann.jspa?annID=1252
Also, I've never tried syncing 700GB using s3cmd sync, much less into
each of the 8 regions...
I'll look into what tool we use to sync our repos into S3.
--Max
_______________________________________________
advisory-board mailing list
advisory-board@xxxxxxxxxxxxxxxxxxxxxxx
https://admin.fedoraproject.org/mailman/listinfo/advisory-board
