Horton

Re: Add jsonlog log_destination for JSON server logs

Nathan Bossart <bossartn@amazon.com>

View thread
From: "Bossart, Nathan" <bossartn@amazon.com>
To: Michael Paquier <michael@paquier.xyz>, Andrew Dunstan <andrew@dunslane.net>
Cc: Andres Freund <andres@anarazel.de>, Sehrope Sarkuni <sehrope@jackdb.com>, Pg Hackers <pgsql-hackers@postgresql.org>, "david@fetter.org" <david@fetter.org>
Date: 2022-01-11T20:34:26Z
Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Introduce log_destination=jsonlog

  2. Refactor set of routines specific to elog.c

  3. Fix compilation warning in syslogger.c

  4. Refactor per-destination file rotation in logging collector

  5. Refactor output file handling when forking syslogger under EXEC_BACKEND

  6. Add regression tests for csvlog with the logging collector

  7. Refactor the syslogger pipe protocol to use a bitmask for its options

On 1/10/22, 4:51 AM, "Michael Paquier" <michael@paquier.xyz> wrote:
> The issue comes from an incorrect change in syslogger_parseArgs()
> where I missed that the incrementation of argv by 3 has no need to be
> changed.  A build with -DEXEC_BACKEND is enough to show the failure,
> which caused a crash when starting up the syslogger because of a NULL
> pointer dereference.  The attached v9 should be enough to switch the
> CF bot to green.

I've been looking at the latest patch set intermittently and playing
around with jsonlog a little.  It seems to work well, and I don't have
any significant comments about the code.  0001 and 0002 seem
straightforward and uncontroversial.  IIUC 0003 simply introduces
jsonlog using the existing framework.

I wonder if we should consider tracking each log destination as a set
of function pointers.  The main logging code would just loop through
the enabled log destinations and use these functions, and it otherwise
would be completely detached (i.e., no "if jsonlog" blocks).  This
might open up the ability to define custom log destinations via
modules, too.  However, I don't know if there's any real demand for
something like this, and it should probably be done separately from
introducing jsonlog, anyway.

Nathan