Create S3 bucket in known region (default to us-east-1) #5381
Conversation
|
Regulatory reasons or compliance may compel one to only store data in a specific region. S3 enables this, because even though S3-buckets are globally named, the underlying storage is per-region. Therefore it would be great if the possibility remains to specify it. Would you please consider this? Using us-east-1 as a default is totally fine, of course. |
|
Justin, would you be opposed to adding it as another config parameter in config-default.sh and config-test.sh? |
|
Not at all... working on it now! LMK if any objections to AWS_S3_REGION. (I think it needs to be a different default from the compute, because of the global namespacing) |
|
AWS_S3_REGION sounds good to me |
|
This is conflicting with #5376. Whichever merges first, I'll sort out the gitmess :-) |
|
|
|
Well neither one is currently mergable according to github, so you may have to sort out two git messes :) |
justinsb
force-pushed
the
aws_s3_in_us_east
branch
2 times, most recently
from
1c29082 to
2435c96
Compare
Buckets are globally named, so we want a single bucket for all our regions.
justinsb
force-pushed
the
aws_s3_in_us_east
branch
from
2435c96 to
2413d25
Compare
justinsb
changed the title
|
I'm not sure what causes a notification in github, so I'm adding a comment in case this has just been overlooked. I have rebased so this should now be mergeable. |
|
Yeah, I'm sorry, Github either doesn't notify reviewers of new commits or my account has been misconfigured for as long as I can remember. Thanks for pinging me. |
Create S3 bucket in known region (default to us-east-1)
Buckets are globally named, so we want a single bucket for all our regions.
us-east-1 is the canonical choice.