Skip to content
This repository has been archived by the owner on Mar 24, 2020. It is now read-only.

HSM use case #25

Open
justinlittman opened this issue Jun 2, 2011 · 4 comments
Open

HSM use case #25

justinlittman opened this issue Jun 2, 2011 · 4 comments

Comments

@justinlittman
Copy link

Much of the content at LC resides on an HSM. To get access to the content, it must be staged from tape to spinning disk. This will take an indeterminate amount of time. A mechanism should be specified for the server to indicate to client that content is not immediately available, but will be at some point in the future, and for the client to check when the content is available for retrieval.

@acdha
Copy link
Owner

acdha commented Jun 2, 2011

Very good topic - do you think this paragraph in the HTTP citizenship section covers it? I was debating whether 504 Gateway Timeout would be more appropriate but in either case does it seem like this would be correct to simply describe in prose?

Clients MUST honor HTTP 503 Service Unavailable responses using a provided Retry-After header
or using exponential back-off if Retry-After is not provided.

@justinlittman
Copy link
Author

Given how our HSM works, I would have 2 choices: only stage the files in a bag which are requested, which places a burden on the client or stage all files when a single file is requested, which is expensive if only a subset of file is required. It would be preferable for a client to be able to request that an entire bag be staged; otherwise per file staging would be used.


From: acdha [[email protected]]
Sent: Thursday, June 02, 2011 4:17 PM
To: Littman, Justin
Subject: Re: [restful-bag-server] HSM use case (#25)

Very good topic - do you think this paragraph in the HTTP citizenship section covers it? I was debating whether 504 Gateway Timeout would be more appropriate but in either case does it seem like this would be correct to simply describe in prose?

Clients MUST honor HTTP 503 Service Unavailable responses using a provided Retry-After header
or using exponential back-off if Retry-After is not provided.

Reply to this email directly or view it on GitHub:
#25 (comment)

@mgschultz
Copy link

Justin - does your use case require some logic within the spec to accommodate your staging requirements, or do we just need to encourage proper configurations at the server side? Sounds like Chris's entry under our Good HTTP Citizenship doesn't fully address your concern.

@justinlittman
Copy link
Author

This mechanism does't allow the server to distinguish between requests for individual files and requests for the entire bag. (So the server knows what to stage.)

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants