Horton

Re: Add support for logging the current role

Stephen Frost <sfrost@snowman.net>

View thread
From: Stephen Frost <sfrost@snowman.net>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Robert Haas <robertmhaas@gmail.com>, PostgreSQL-development <pgsql-hackers@postgresql.org>
Date: 2011-01-13T01:21:54Z
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. Add support for an application_name parameter, which is displayed in

  2. Make CSV column ordering a bit more logical.

  3. Extend the format of CSV logs to include the additional information supplied

  4. Add virtual transaction IDs to CSVLOG output, so that messages coming from

  5. Provide for logfiles in machine readable CSV format. In consequence, rename

* Tom Lane (tgl@sss.pgh.pa.us) wrote:
> Stephen Frost <sfrost@snowman.net> writes:
> > It might be possible to do and answer that specific question- but what
> > about the obvious next question: which role was this command run with?
> > iow, if I log dml, how do I know what the role was when the dml
> > statement was run?  ie- why was this command allowed?
> 
> I'm less than excited about that argument because it's after the fact
> --- if you needed to know the information, you probably didn't have
> log_line_prefix set correctly, even assuming you had adequate logging
> otherwise.  And logging an OID just seems too ugly to live.

Erm, really?  Ok, fine, maybe you didn't have log_line_prefix set
correctly the first time you needed the information, but after you
discover that you *don't know*, you're going to be looking for an option
to let you get that information for the future.  I would also suggest
that more experienced admins are going to have a default log_line_prefix
that they install on new systems they set up (I know I do...), and I'd
be suprised if knowing the role that a command is actually run as wasn't
popular among that set.

I don't like logging an OID either.

> Another little problem with the quick and dirty solution is that stuff
> that's important enough to warrant a log_line_prefix escape is generally
> thought to be important enough to warrant inclusion in CSV logs.  That
> would imply adding a column and taking the resultant compatibility hit.

I'd be more than happy to add support for this to the CSV logs.  I agree
that it'd make sense to do.  I think we need to solve the bigger problem
of OID vs. rolename vs. lookups from elog first though.

	Thanks,

		Stephen