Closed scaleoutsean closed 2 years ago
Hi @scaleoutsean, apologies for the delayed response. I've opened https://jira.ngage.netapp.com/browse/DOC-3978 to track the improvements you suggested. Can you expand on your S3 suggestion?
For your bucket cloning questions, Here are some answers from product management: -=-=-=-=- In short, no. A bucket is a construct within a single hyperscaler project and is not intended to be portable.
You CAN clone a k8s application from one hyperscaler to another or between regions in a single hyperscaler today with ACS. What you actually have is a bucket in each hyperscaler and the data gets replicated by ACS from one to the other. But that is a single backup within a bucket and not a wholesale clone of the bucket with all of its backups inside. -=-=-=-=-
If this is helpful, I can make the above points more clear in the ACS documentation.
Regarding S3, yes, they do because it's a kind of universal short query for on-prem object storage which normally supports AWS S3-like API calls. Another reason is some want to find if ACC backup target can be replicated to AWS S3 for DR purposes and used to restore backups either directly from that target (once ACS can "import" such repos) or replicate back to on-prem in the case on-prem object store needs to be restored. (I am not asking for these features, but people attempt to search for this information)
Thanks - I've added S3 and related keywords to that topic, and also added the following info to the buckets introduction:
"You can not clone a bucket from one cloud provider to another cloud provider. To migrate data from one cloud provider to another, clone the associated application instead."
If you approve, I'll make this change public. Thanks!
Looks good to me
Page: Manage buckets
1) Add "S3" (edit: blob, etc.) somewhere in the document to improve searchability
2) Explain if it's possible (or not?) to re-assign or clone buckets from one ACC/ACS front end to another and whether that works across hyperscalers. For example:
Can I clone a bucket from hyperscaler A to a bucket at hyperscaler B and use it by ACS instance running there? Or can the clone buckets be accessed only by Astra Control instance that created them? Of course, the reader can find out experimentally, but it's time consuming because it'd require multiple AC instances in different hyperscalers, so it's better to have these details explained