Shibboleth Developers

[Chad La Joie <>]

The IdP performs an HTTP HEAD request and only if the last update time 
is newer does it perform a GET.  It tries to be a good netizen.  :)

Ian Young wrote:
> This is a broad question about both the 2.0 IdP and the 2.0 SP.
>
> With 1.3, we all made cron jobs to go fetch and verify metadata 
> periodically.  In 2.0, both the IdP and SP support performing periodic 
> fetch and verify themselves, along with cacheing a local copy.  This is 
> all good.
>
> My question is whether the new code always fetches the full file, or 
> whether they try and optimise bandwidth by not fetching the body if the 
> file hasn't changed since the last fetch?
>
> Obviously if the new code did do this, we might be more sanguine about 
> telling people to poll more frequently.
>
>     -- Ian
>
> P.S. I don't think that you'd need to preserve "last fetch time" over a 
> restart; it would be perfectly fine for the code to always fetch when it 
> was restarted.

--
SWITCH
Serving Swiss Universities
--------------------------
Chad La Joie, Software Engineer, Security
Werdstrasse 2, P.O. Box, 8021 Zürich, Switzerland
phone +41 44 268 15 75, fax +41 44 268 15 68
,
 http://www.switch.ch