Re: Call for Adoption: Proxy Status

[Alex Rousskov <[e-mail]>]

On 4/22/19 6:34 PM, Mark Nottingham wrote:
>> On 22 Apr 2019, at 2:50 am, Alex Rousskov wrote:
>> On 4/10/19 6:24 PM, Tommy Pauly wrote:
>>> https://tools.ietf.org/html/draft-nottingham-proxy-status-00

>> I believe the scope of the document should be enlarged from the current
>> "error details in the proxy-generated response" to something like "proxy
>> status(es) when handling the message". After that, it should be adopted.

> Agreed; Piotr and I have already been discussing that. That said, we
> should be careful to understand the delineation between this and the
> Cache header.

I doubt we need two header fields sharing the same goal of reporting
what happened at the proxy. One status header field is enough AFAICT.
Any caching-related statuses are a subset of proxy statuses. The
"universal" header must have a list syntax, but that syntax is required
for each of the two header fields anyway.

If the Cache header has already been standardized, this draft can
deprecate its early limited usage in favor of a "universal" Proxy-Status.

Alex.



>> IMHO, there is no good reason to restrict a generic "Proxy-Status"
>> mechanism to proxy-generated errors, especially since existing proxies
>> already use similar mechanisms to relay their state when forwarding
>> messages (in addition to generating error responses). Typical uses
>> include relaying caching state (initial lookup outcome, refresh
>> activity, etc.) and proxy-specific transaction IDs (for correlating
>> messages with proxy logs).