Re: Patch proposal: New hooks in the connection path - Mailing list pgsql-hackers

From Bharath Rupireddy
Subject Re: Patch proposal: New hooks in the connection path
Date
Msg-id CALj2ACVBcOTkpCDqKiZX=+L+8aigc-+12-eQuFSALtcYR6Z=DQ@mail.gmail.com
Whole thread Raw
In response to Re: Patch proposal: New hooks in the connection path  ("Drouvot, Bertrand" <bdrouvot@amazon.com>)
Responses Re: Patch proposal: New hooks in the connection path
List pgsql-hackers
On Mon, Jul 4, 2022 at 6:23 PM Drouvot, Bertrand <bdrouvot@amazon.com> wrote:
>
> Hi,
>
> On 7/2/22 1:00 AM, Nathan Bossart wrote:
> > Could we model this after fmgr_hook?  The first argument in that hook
> > indicates where it is being called from.  This doesn't alleviate the need
> > for several calls to the hook in the authentication logic, but extension
> > authors would only need to define one hook.
>
> I like the idea and indeed fmgr.h looks a good place to model it.
>
> Attached a new patch version doing so.

Thanks for the patch. Can we think of enhancing
ClientAuthentication_hook_type itself i.e. make it a generic hook for
all sorts of authentication metrics, info etc. with the type parameter
embedded to it instead of new hook FailedConnection_hook?We can either
add a new parameter for the "event" (the existing
ClientAuthentication_hook_type implementers will have problems), or
embed/multiplex the "event" info to existing Port structure or status
variable (macro or enum) (existing implementers will not have
compatibility problems).  IMO, this looks cleaner going forward.

On the v2 patch:

1. Why do we need to place the hook and structures in fmgr.h? Why not in auth.h?

2. Timeout Handler is a signal handler, called as part of SIGALRM
signal handler, most of the times, signal handlers ought to be doing
small things, now that we are handing off the control to hook, which
can do any long running work (writing to a remote storage, file,
aggregate etc.), I don't think it's the right thing to do here.
 static void
 StartupPacketTimeoutHandler(void)
 {
+ if (FailedConnection_hook)
+ (*FailedConnection_hook) (FCET_STARTUP_PACKET_TIMEOUT, MyProcPort);
+ ereport(COMMERROR,
+ (errcode(ERRCODE_PROTOCOL_VIOLATION),
+ errmsg("timeout while processing startup packet")));

3. On "not letting these hooks (ClientAuthentication_hook_type or
FailedConnection_hook_type) expose sensitive info via Port structure"
- it seems like the Port structure has sensitive info like HbaLine,
host address, username, etc. but that's what it is so I think we are
okay with the structure as-is.

Regards,
Bharath Rupireddy.



pgsql-hackers by date:

Previous
From: Amit Langote
Date:
Subject: Re: doc phrase: "inheritance child"
Next
From: Noah Misch
Date:
Subject: Re: typos