- My Forums
- Tiger Rant
- LSU Recruiting
- SEC Rant
- Saints Talk
- Pelicans Talk
- More Sports Board
- Fantasy Sports
- Golf Board
- Soccer Board
- O-T Lounge
- Tech Board
- Home/Garden Board
- Outdoor Board
- Health/Fitness Board
- Movie/TV Board
- Book Board
- Music Board
- Political Talk
- Money Talk
- Fark Board
- Gaming Board
- Travel Board
- Food/Drink Board
- Ticket Exchange
- TD Help Board
Customize My Forums- View All Forums
- Show Left Links
- Topic Sort Options
- Trending Topics
- Recent Topics
- Active Topics
Started By
Message
S3 baws; anyone familiar with their CDN & Edge configuration & management?
Posted on 9/6/23 at 4:04 pm
Posted on 9/6/23 at 4:04 pm
once upon a time i was in charge of integrating with akamai and so almost 20 years later everything is jaded by that model but we are pushing parts of our app to AWS and the TERRABYTES of file storage actively used by our customers is a hot topic.
i'll try to keep my objectives simple:
- we currently host hundreds of thousands of files on our own hardware/web file-delivery servers
- would like to use that going forward as the seed/pull-from location to populate the CDN/edge servers. is that possible? many of those files may never be requested by end users so it seems a waste to copy the entire server up to S3
- speaking of that, we have a process where hundreds or even thousands of those files are updated daily. S3 needs to be able to pull those; in the past we set a TTL of 24 hours or something similar to ensure stale files were removed promptly
- we also had a management tool where we could flush a given directory should an outdated file or problem arise where we needed to immediately remove access.
does any of that make sense?
here's the tl;dr
- can S3 storage match an existing file/directory structure hosted on our hardware and do we have to push the entire file server to S3 before it can manage and push those files to the Edge servers?
i'll try to keep my objectives simple:
- we currently host hundreds of thousands of files on our own hardware/web file-delivery servers
- would like to use that going forward as the seed/pull-from location to populate the CDN/edge servers. is that possible? many of those files may never be requested by end users so it seems a waste to copy the entire server up to S3
- speaking of that, we have a process where hundreds or even thousands of those files are updated daily. S3 needs to be able to pull those; in the past we set a TTL of 24 hours or something similar to ensure stale files were removed promptly
- we also had a management tool where we could flush a given directory should an outdated file or problem arise where we needed to immediately remove access.
does any of that make sense?
here's the tl;dr
- can S3 storage match an existing file/directory structure hosted on our hardware and do we have to push the entire file server to S3 before it can manage and push those files to the Edge servers?
This post was edited on 9/6/23 at 4:06 pm
Posted on 9/6/23 at 5:37 pm to CAD703X
I am just an AWS CCP, so I'm spit balling. You are charging customers for the TB they're using right? That makes their data migration their cost decision, not yours. If they are willing to pay to store TB of data on S3, they get to store TB of data on S3.
Doesn't AWS already have tools to move stuff from S3 to Glacier in a relatively automated fashion? You could push non stale files into S3, and push stale files into Glacier, just in case. I don't know about the guts of Storage Gateway, but that could be an answer.
This Stackoverflow question (which wasn't answered, from 8 years ago) might give you some threads to pull. It's over my head, but you might recognize something useful.
Doesn't AWS already have tools to move stuff from S3 to Glacier in a relatively automated fashion? You could push non stale files into S3, and push stale files into Glacier, just in case. I don't know about the guts of Storage Gateway, but that could be an answer.
This Stackoverflow question (which wasn't answered, from 8 years ago) might give you some threads to pull. It's over my head, but you might recognize something useful.
Posted on 9/6/23 at 6:10 pm to CAD703X
Been almost a year since i was up to date on AWS stuff, but sounds like the perfect use case. Your hardware as storage gateway and pushed up to aws, set retention or even tiered for frequent/infrequent access. All which can be automated.
Popular
Back to top
