[ckan-dev] CKAN as an API Proxy

[Andy Collins]

Thank you, Stefan. 

We will be posting publicly available data, but there are still a few reasons we would like authentication. Mostly it’s a mechanism to support communication with our API consumers (we’d collect contact information from each consumer). Also we anticipate developers using the APIs in applications and would like to be able to throttle requests if that became necessary. Finally it does leave the door open to publishing more sensitive data.

If we decided to use CKAN, I guess we’ll need to build something…or more likely modify an existing API proxy. If we did that do you think there’d be any value from sticking with python? We are debating about python verses a node solution.

Thanks again for you quick and helpful response.

—Andy.

On Jan 9, 2015, at 6:09 AM, <ckan-dev-request at lists.okfn.org> <ckan-dev-request at lists.okfn.org> wrote:

> Date: Fri, 9 Jan 2015 13:09:05 +0100
> From: Stefan Oderbolz <stefan.oderbolz at liip.ch>
> To: CKAN Development Discussions <ckan-dev at lists.okfn.org>
> Subject: Re: [ckan-dev] CKAN as an API Proxy
> Message-ID:
>        <CAARrFXcpxt-j2fL35hFKGbNQbE9E0VTAx8TeMMn1VsOBHOn9bA at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
> 
> Hi Andy,
> 
> I quite sure this does not yet exist in the context of CKAN, but there are
> a lot of similar solutions for API proxies.
> Your use case sounds like you need an external API proxy, which handles
> requests by the "outside world". The authentication could be done via CKAN,
> but I'm not sure why you need it anyway. Is this a private instance of
> CKAN? If so, it makes sense, if the data is open, I wouldn't implement the
> authentication part.
> 
> Apart from the API proxy you want to extend the data model to store two
> URLs, one for the proxy and one for the "real" API. This should be doable.
> The API proxy can then use the information saved on the CKAN resource to
> route the requests. I personally would use something like hapi.js to build
> the proxy.
> 
> - Stefan
> 
> 
> On Thu, Jan 8, 2015 at 10:31 PM, Andy Collins <Andy.Collins at tn.gov> wrote:
> 
>> 
>> 
>> We would like to use CKAN as an API proxy or to find a proxy that would
>> work well with CKAN.
>> 
>> These APIs will be live (?real time?) data feeds and not static datasets.
>> 
>> We have the following desires for such a proxy:
>> 1. We do not want to expose our back-end source APIs to the Internet. All
>> APIs would be requested via the proxy.
>> 2. We would like to take advantage of CKAN's user management/api keys.
>> Ideally each request to the API could be authenticated via CKAN. This may
>> be by using CKAN's api keys or via oauth.
>> 3. We would like a mechanism for managing both the source API URL and the
>> proxy API URL. Ideally there would be one place to create a resource where
>> each URL could be specified.
>> 
>> 
>> Attached is an image of a rough diagram of the kind of system we are
>> considering building, but we?re hoping someone has blazed this trail
>> already. Does anyone have any ideas or suggestions for solutions we should
>> explore?
>> 
>> Thank you.
>> 
>> ?Andy.
>> 
>> 
>> _______________________________________________
>> ckan-dev mailing list
>> ckan-dev at lists.okfn.org
>> https://lists.okfn.org/mailman/listinfo/ckan-dev
>> Unsubscribe: https://lists.okfn.org/mailman/options/ckan-dev
>> 
>> 
> 
> 
> --
> Liip AG  // Limmatstrasse 183 //  CH-8005 Z?rich
> Tel +41 43 500 39 80 // GnuPG 0x7B588C67 // www.liip.ch
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.okfn.org/pipermail/ckan-dev/attachments/20150109/9d6943a7/attachment.html>
> -------------- next part --------------
> A non-text attachment was scrubbed...
> Name: CKAN_Proxy.png
> Type: image/png
> Size: 63004 bytes
> Desc: not available
> URL: <http://lists.okfn.org/pipermail/ckan-dev/attachments/20150109/9d6943a7/attachment.png>
> 
> ------------------------------
> 
> Subject: Digest Footer
> 
> _______________________________________________
> ckan-dev mailing list
> ckan-dev at lists.okfn.org
> https://lists.okfn.org/mailman/listinfo/ckan-dev
> Unsubscribe: https://lists.okfn.org/mailman/options/ckan-dev
> 
> 
> ------------------------------
> 
> End of ckan-dev Digest, Vol 51, Issue 10
> ****************************************