Tracing edupersontargetedid to usernames

[Steve Glover]

On 27/05/14 15:56, Peter Schober wrote:

>> Or am I missing something horribly obvious?

(clearly, I was)

> Not sure what you're saying, the aacli doesn't take ePTId as an
> input. Are you suggesting to loop over all userids with the aacli?

I had been. I didn't consider the issues involved in scaling up from a 
test IdP with less than ten "users"

> That would work but I'm guessing will be /much/ slower than simply
> finding and re-implementing the algorithm used in the generation of
> the values and doing the looping outside of repeated JVM startups and
> teardowns for every single subject.

Hadn't realised there was quite so much overhead - I just went to 
another screen and left the script running (once I'd found out about the 
whole missing servlet-api.jar thing and upgraded to 2.4.0).

But yeah, on checking, running the aacli.sh script eight times took 
2m48s - scaling to any useful number of users to generate the epTID 
would be silly (that said, "simply finding and re-implementing the 
algorithm" would probably take me an even more ridiculous amount of time).

It occurs to me that a tool to generate ePTIDs for all an IdP's users 
might be useful for just this sort of occasion

Steve



> -peter
> --
> To unsubscribe from this list send an email to users-unsubscribe at shibboleth.net
>
>
>


-- 
Steve Glover: SDSS, EDINA, Causewayside House, 160 Causewayside EH9 1PR
e:steve.glover at ed.ac.uk t:0131 650 2908 f:0131 650 3308 m:07961 446 902

The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.