issuers
6 views
Skip to first unread message
Withers, Alexander
Jun 14, 2019, 3:17:02 PM6/14/19
to SciTokens Discussion, Gao, You
Hello All,
Alex G and I were discussing some of the next steps for the apache-scitokens module. Alex G was wondering if there are other issuers aside from demo.scitokens.org? Additionally, he has some testing requirements but I’ll let him fill in the technical details for that.
Thanks,
Alex W.
Jeff Gaynor
Jun 14, 2019, 3:34:07 PM6/14/19
to dis...@scitokens.org
There is the demo at
https://surge.ncsa.illinois.edu/scitokens-client/ which will issue
SciTokens. He can use the demo as it is set up and if he wants to
have his own client (which will then issue SciTokens securely for
whatever paths and services he actually needs) then we can do
that.
Jeff
--
You received this message because you are subscribed to the Google Groups "SciTokens Discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to discuss+u...@scitokens.org.
To view this discussion on the web visit https://groups.google.com/a/scitokens.org/d/msgid/discuss/DM6PR11MB2714987658AFCE8CF278095EEDEE0%40DM6PR11MB2714.namprd11.prod.outlook.com.
Brian Bockelman
Jun 15, 2019, 12:09:26 PM6/15/19
to Withers, Alexander, SciTokens Discussion, Gao, You
Paul Millar
Jun 17, 2019, 4:53:47 AM6/17/19
to dis...@scitokens.org
Hi all,
You might like to try an HTTP third-party copy that uses SciTokens.
To do this, you can use prometheus.desy.de, a dCache test instance that
supports DTEAM and CMS SciTokens via the two OPs Brian mentioned.
An example of such a demo would involve making a request to dCache to
copy a file from an Apache server, handing over a SciToken to authorise
dCache to read the file, and using a SciToken (possible the same one) to
authorise the third-party copy.
dCache would make a GET request with the SciToken to fetch the data.
For the Apache server, this would appear like a regular HTTP GET request
(with a SciToken), so it shouldn't require any changes.
The opposite direction is also supported: data within dCache may be
written to the Apache server, using an HTTP PUT request.
Cheers,
Paul.
On 15/06/2019 18:09, Brian Bockelman wrote:
> Hi,
>
> How about:
> - scitokens.org/dteam <http://scitokens.org/dteam>
> - cmsweb.cern.ch <http://cmsweb.cern.ch>
> <https://groups.google.com/a/scitokens.org/d/msgid/discuss/43EE62EA-FAA1-43E3-8473-0C127C2F95D5%40gmail.com?utm_medium=email&utm_source=footer>.
You might like to try an HTTP third-party copy that uses SciTokens.
To do this, you can use prometheus.desy.de, a dCache test instance that
supports DTEAM and CMS SciTokens via the two OPs Brian mentioned.
An example of such a demo would involve making a request to dCache to
copy a file from an Apache server, handing over a SciToken to authorise
dCache to read the file, and using a SciToken (possible the same one) to
authorise the third-party copy.
dCache would make a GET request with the SciToken to fetch the data.
For the Apache server, this would appear like a regular HTTP GET request
(with a SciToken), so it shouldn't require any changes.
The opposite direction is also supported: data within dCache may be
written to the Apache server, using an HTTP PUT request.
Cheers,
Paul.
On 15/06/2019 18:09, Brian Bockelman wrote:
> Hi,
>
> How about:
> - cmsweb.cern.ch <http://cmsweb.cern.ch>
>
> Brian
>
> Sent from my iPhone
>
> On Jun 14, 2019, at 2:16 PM, Withers, Alexander <ale...@illinois.edu
> Brian
>
> Sent from my iPhone
>
> On Jun 14, 2019, at 2:16 PM, Withers, Alexander <ale...@illinois.edu
> <mailto:ale...@illinois.edu>> wrote:
>
>> Hello All,
>>
>> Alex G and I were discussing some of the next steps for the
>> apache-scitokens module. Alex G was wondering if there are other
>> issuers aside from demo.scitokens.org <http://demo.scitokens.org>?
>
>> Hello All,
>>
>> Alex G and I were discussing some of the next steps for the
>> apache-scitokens module. Alex G was wondering if there are other
>> Additionally, he has some testing requirements but I’ll let him fill
>> in the technical details for that.
>>
>> Thanks,
>>
>> Alex W.
>>
>> --
>> You received this message because you are subscribed to the Google
>> Groups "SciTokens Discussion" group.
>> To unsubscribe from this group and stop receiving emails from it, send
>> an email to discuss+u...@scitokens.org
>> <mailto:discuss+u...@scitokens.org>.
>> in the technical details for that.
>>
>> Thanks,
>>
>> Alex W.
>>
>> --
>> You received this message because you are subscribed to the Google
>> Groups "SciTokens Discussion" group.
>> To unsubscribe from this group and stop receiving emails from it, send
>> an email to discuss+u...@scitokens.org
>> To view this discussion on the web visit
>> https://groups.google.com/a/scitokens.org/d/msgid/discuss/DM6PR11MB2714987658AFCE8CF278095EEDEE0%40DM6PR11MB2714.namprd11.prod.outlook.com
>> <https://groups.google.com/a/scitokens.org/d/msgid/discuss/DM6PR11MB2714987658AFCE8CF278095EEDEE0%40DM6PR11MB2714.namprd11.prod.outlook.com?utm_medium=email&utm_source=footer>.
>> https://groups.google.com/a/scitokens.org/d/msgid/discuss/DM6PR11MB2714987658AFCE8CF278095EEDEE0%40DM6PR11MB2714.namprd11.prod.outlook.com
>
> --
> You received this message because you are subscribed to the Google
> Groups "SciTokens Discussion" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to discuss+u...@scitokens.org
> <mailto:discuss+u...@scitokens.org>.
> --
> You received this message because you are subscribed to the Google
> Groups "SciTokens Discussion" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to discuss+u...@scitokens.org
> To view this discussion on the web visit
> https://groups.google.com/a/scitokens.org/d/msgid/discuss/43EE62EA-FAA1-43E3-8473-0C127C2F95D5%40gmail.com
> <https://groups.google.com/a/scitokens.org/d/msgid/discuss/43EE62EA-FAA1-43E3-8473-0C127C2F95D5%40gmail.com?utm_medium=email&utm_source=footer>.
Brian Bockelman
Jun 17, 2019, 8:37:19 AM6/17/19
to Paul Millar, SciTokens Discussion
You might also appreciate the following repository / PR:
https://github.com/paulmillar/http-tpc-utils/pull/9
That allows one to test various HTTP third-party-copy mechanisms
solely using SciTokens.
(Paul - I see that the PR has developed a merge conflict... I can fix
that, but any chance you'll have time to merge in the near future?)
Brian
> To unsubscribe from this group and stop receiving emails from it, send an email to discuss+u...@scitokens.org.
> To view this discussion on the web visit https://groups.google.com/a/scitokens.org/d/msgid/discuss/360e4e42-81ef-27ae-9116-10bbacd48804%40desy.de.
https://github.com/paulmillar/http-tpc-utils/pull/9
That allows one to test various HTTP third-party-copy mechanisms
solely using SciTokens.
(Paul - I see that the PR has developed a merge conflict... I can fix
that, but any chance you'll have time to merge in the near future?)
Brian
> To view this discussion on the web visit https://groups.google.com/a/scitokens.org/d/msgid/discuss/360e4e42-81ef-27ae-9116-10bbacd48804%40desy.de.
Paul Millar
Jun 17, 2019, 9:49:57 AM6/17/19
to Brian Bockelman, SciTokens Discussion
Hi everyone!
On 17/06/2019 14:37, Brian Bockelman wrote:
> You might also appreciate the following repository / PR:
>
> https://github.com/paulmillar/http-tpc-utils/pull/9
>
> That allows one to test various HTTP third-party-copy mechanisms
> solely using SciTokens.
This is good (and appreciated) but might be somewhat orthogonal, since
the purpose is to test an endpoint's support for the HTTP-TPC protocol.
Adding HTTP-TPC protocol support as an Apache module would be great, but
perhaps not the focus of Alex G. and W.'s work. (Right?).
As it happens, it's very easy to trigger a third-party copy with curl.
If let me know your endpoint, I can send you the command-line to trigger
this with prometheus.desy.de
> (Paul - I see that the PR has developed a merge conflict... I can fix
> that, but any chance you'll have time to merge in the near future?)
Yes, absolutely!
I noticed the merge conflict, too, and send an email to request you
update the pull-request. (I guess that was to one of the wrong email
addresses ;-)
I'm just testing some patches to add GOCDB support. So I suggest you
hold off until those are committed, which should be later on today.
Cheers,
Paul.
On 17/06/2019 14:37, Brian Bockelman wrote:
> You might also appreciate the following repository / PR:
>
> https://github.com/paulmillar/http-tpc-utils/pull/9
>
> That allows one to test various HTTP third-party-copy mechanisms
> solely using SciTokens.
the purpose is to test an endpoint's support for the HTTP-TPC protocol.
Adding HTTP-TPC protocol support as an Apache module would be great, but
perhaps not the focus of Alex G. and W.'s work. (Right?).
As it happens, it's very easy to trigger a third-party copy with curl.
If let me know your endpoint, I can send you the command-line to trigger
this with prometheus.desy.de
> (Paul - I see that the PR has developed a merge conflict... I can fix
> that, but any chance you'll have time to merge in the near future?)
I noticed the merge conflict, too, and send an email to request you
update the pull-request. (I guess that was to one of the wrong email
addresses ;-)
I'm just testing some patches to add GOCDB support. So I suggest you
hold off until those are committed, which should be later on today.
Cheers,
Paul.
Reply all
Reply to author
Forward
0 new messages