You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@trafficserver.apache.org by Miles Libbey <ml...@apache.org> on 2020/01/16 22:41:28 UTC
Proposal for 9.x -- change X-Debug plugin's X-Cache order
Hi folks-
In https://github.com/apache/trafficserver/pull/6328 we document the
X-Debug plugin's behavior with multiple proxies. Unfortunately, it's
done in the opposite order of the Via header (which is append, based
on https://tools.ietf.org/html/rfc7230#section-5.7.1). This is
confusing -- we can have 2 headers next to each other with similar
data, but, needs to be read in the exact opposite order.
Proposal is to standardize the X-Cache header to behave like Via. This
would be an incompatible change, so would need to go into a major
release.
Support? Concerns? Objections?
miles
Re: Proposal for 9.x -- change X-Debug plugin's X-Cache order
Posted by Aaron Canary <ac...@verizonmedia.com>.
+1, I support this proposal.
On Thu, Jan 16, 2020 at 4:41 PM Miles Libbey <ml...@apache.org> wrote:
> Hi folks-
>
> In https://github.com/apache/trafficserver/pull/6328 we document the
> X-Debug plugin's behavior with multiple proxies. Unfortunately, it's
> done in the opposite order of the Via header (which is append, based
> on https://tools.ietf.org/html/rfc7230#section-5.7.1). This is
> confusing -- we can have 2 headers next to each other with similar
> data, but, needs to be read in the exact opposite order.
>
> Proposal is to standardize the X-Cache header to behave like Via. This
> would be an incompatible change, so would need to go into a major
> release.
>
> Support? Concerns? Objections?
>
> miles
>
Re: Proposal for 9.x -- change X-Debug plugin's X-Cache order
Posted by Leif Hedstrom <zw...@apache.org>.
> On Jan 16, 2020, at 3:41 PM, Miles Libbey <ml...@apache.org> wrote:
>
> Hi folks-
>
> In https://github.com/apache/trafficserver/pull/6328 we document the
> X-Debug plugin's behavior with multiple proxies. Unfortunately, it's
> done in the opposite order of the Via header (which is append, based
> on https://tools.ietf.org/html/rfc7230#section-5.7.1). This is
> confusing -- we can have 2 headers next to each other with similar
> data, but, needs to be read in the exact opposite order.
>
> Proposal is to standardize the X-Cache header to behave like Via. This
> would be an incompatible change, so would need to go into a major
> release.
Yeh, +1 for me too. I made a PR for review:
https://github.com/apache/trafficserver/pull/6333 <https://github.com/apache/trafficserver/pull/6333>
— leif
Re: Proposal for 9.x -- change X-Debug plugin's X-Cache order
Posted by Aaron Canary <ac...@verizonmedia.com.INVALID>.
+1, I support this proposal.
On Thu, Jan 16, 2020 at 4:41 PM Miles Libbey <ml...@apache.org> wrote:
> Hi folks-
>
> In https://github.com/apache/trafficserver/pull/6328 we document the
> X-Debug plugin's behavior with multiple proxies. Unfortunately, it's
> done in the opposite order of the Via header (which is append, based
> on https://tools.ietf.org/html/rfc7230#section-5.7.1). This is
> confusing -- we can have 2 headers next to each other with similar
> data, but, needs to be read in the exact opposite order.
>
> Proposal is to standardize the X-Cache header to behave like Via. This
> would be an incompatible change, so would need to go into a major
> release.
>
> Support? Concerns? Objections?
>
> miles
>
Re: Proposal for 9.x -- change X-Debug plugin's X-Cache order
Posted by Leif Hedstrom <zw...@apache.org>.
> On Jan 16, 2020, at 3:41 PM, Miles Libbey <ml...@apache.org> wrote:
>
> Hi folks-
>
> In https://github.com/apache/trafficserver/pull/6328 we document the
> X-Debug plugin's behavior with multiple proxies. Unfortunately, it's
> done in the opposite order of the Via header (which is append, based
> on https://tools.ietf.org/html/rfc7230#section-5.7.1). This is
> confusing -- we can have 2 headers next to each other with similar
> data, but, needs to be read in the exact opposite order.
>
> Proposal is to standardize the X-Cache header to behave like Via. This
> would be an incompatible change, so would need to go into a major
> release.
Yeh, +1 for me too. I made a PR for review:
https://github.com/apache/trafficserver/pull/6333 <https://github.com/apache/trafficserver/pull/6333>
— leif