Per-Entity Metadata Working Group

[Nick Roy <>]

Just my personal opinions, take with appropriate grain(s) of salt.

On 9/14/16 10:20 AM, Tom Scavo wrote:
> I still have a number of unanswered questions regarding the
> distribution of per-entity metadata:
>
> 1) Do we need preview-main-fallback servers or is a single production
> server adequate?

From what I understand about TIER's intentions and directions from 
Steve Zoppi (and I agree with them), the long-term intent is to create 
an entire test federation in the TIER QA environment that we could use 
to test out new stuff (including the interaction of TIER components with 
changes to metadata).  This is one reason I'm working with Levvel.io 
right now to build out a Federation Manager CI pipeline in the TIER QA 
environment.  The intent is to have a parallel InCommon metadata service 
(metadata curation, production, aggregation, signing, publication) in 
that environment long-term. Maybe that is what's needed?

> 2) Does the server need to support HTTP Conditional GET?

I don't see why - any change to a _signed_ entity descriptor would 
probably mean the entire signed entity descriptor needs to get 
completely re-retrieved.

> 3) What is the range of permissible validUntil dates on each entity
> descriptor? (we discussed this briefly on today's call)

Leave it at 14 days for now, address making the range smaller in the 
future as it becomes feasible.

> 4) Is there a cacheDuration on each entity descriptor, and if so, what
> is its value?

I think the answer to this depends on how frequently we intend to 
publish new signed metadata, so currently, maybe put it at 2 hours to 
ensure changes are picked up in a relatively timely fashion, but we 
aren't hammering the servers too hard.

> 5) What is our failover strategy?

Multiple CDNs with different hostnames, configured at each metadata 
client, with the requirement to implement some kind of client-based 
failover when deployer-controllable (and fed ops recommended) thresholds 
for latency and timeout are exceeded.

Nick

> I realize there's pressure to wrap up the WG but if anyone cares to
> address any of these questions, I'd be grateful.
>
> Tom