Shib IdP 2.4 metadata refresh bug?

[Brent Putman]

On 4/20/15 7:14 PM, Cantor, Scott wrote:
> Was thinking something more environmental in Java that caused the library 
> to regress, but as you say we apparently added the trap and logging to the 
> old code anyway.

Ah, ok, you mean something like a low-level Java change re: sockets or
Timer behavior or something like that.  Yeah, I guess that's possible.

> Still seems odd to me, there were hardly any reports of this kind of 
> problem for a really long time.

That's true.  Either something relevant did change somewhere, or maybe
it just never got reported.  If someone were restarting their IdP
pretty frequently, they might not see it, if the hung metadata was
still valid up to the restart.