From: | "David G(dot) Johnston" <david(dot)g(dot)johnston(at)gmail(dot)com> |
---|---|
To: | Robert Haas <robertmhaas(at)gmail(dot)com> |
Cc: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, Satyanarayana Narlapuram <Satyanarayana(dot)Narlapuram(at)microsoft(dot)com>, "pgsql-hackers(at)postgresql(dot)org" <pgsql-hackers(at)postgresql(dot)org> |
Subject: | Re: Adding connection id in the startup message |
Date: | 2017-06-21 19:45:57 |
Message-ID: | CAKFQuwa0Ao=vikg5n0GgD6ecVA+rM+5JwPg1PmhPeYhgiv+_LA@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On Wed, Jun 21, 2017 at 12:15 PM, Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
> I think the problem is real,
> but I'm not sure that this is the best solution. On the other hand,
> I'm also not entirely sure I understand the proposal yet.
Given the problems with changing the protocol it does seem like
generalizing away from "connection id" to "arbitrary payload" would be
useful.
Like, add a new "attachment map" area where any application-aware node
that encounters the message can attach data. If some node further
downstream sends a response message the contents of the attachment map
would be sent back as-is.
The advantage of "connection id" is that the messages are still of
fixed size whereas the more flexible arrangement would necessitate
that message sizes vary. In return middleware gets a place to store
session information that can be used more broadly than a simple
externally generated connection id.
David J.
From | Date | Subject | |
---|---|---|---|
Next Message | Robert Haas | 2017-06-21 19:47:22 | Re: Pluggable storage |
Previous Message | Peter Eisentraut | 2017-06-21 19:44:12 | Re: Missing comment for ResultRelInfo in execnodes.h |