Re: [HACKERS] Cutting initdb's runtime (Perl question embedded) - Mailing list pgsql-hackers

From Tom Lane
Subject Re: [HACKERS] Cutting initdb's runtime (Perl question embedded)
Date
Msg-id 8873.1525881373@sss.pgh.pa.us
Whole thread Raw
In response to Re: [HACKERS] Cutting initdb's runtime (Perl question embedded)  (Alvaro Herrera <alvherre@2ndquadrant.com>)
Responses Re: [HACKERS] Cutting initdb's runtime (Perl question embedded)  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
Alvaro Herrera <alvherre@2ndquadrant.com> writes:
> Tom Lane wrote:
>> Robert Haas <robertmhaas@gmail.com> writes:
>>> Who says we need a portable way?  If we had something that worked on
>>> Linux and macOS, it would cover most developer environments.  I wonder
>>> if readlink("/etc/localtime", buf, sz) might be a viable approach.

>> I wondered about that, but I'm afraid it's often a hardlink not a
>> symlink.  Still, we could try it.

> In Debian systems, it's a symlink.  Apparently in RHEL6 and older it's a
> copy or hardlink, and the file /etc/sysconfig/clock contains a ZONE
> variable that points to the right zone.

Yeah, on my RHEL6 box,

$ ls -l /etc/localtime 
-rw-r--r--. 1 root root 3519 May  4  2010 /etc/localtime

The mod date's a little astonishing, considering this system was built
in 2013.  It is *not* a hardlink, or even an exact match, to anything
under /usr/share/zoneinfo, though perhaps it was originally.  Also:

$ cat /etc/sysconfig/clock
# The time zone of the system is defined by the contents of /etc/localtime.
# This file is only for evaluation by system-config-date, do not rely on its
# contents elsewhere.
ZONE="America/New York"

I'm inclined to respect the comment, especially since I see they are not
spelling the zone name canonically anyway (note space not underscore);
so looking at this wouldn't work without some ill-defined heuristics.

However, more modern Red Hat systems seem to have /etc/localtime as
a symlink, so checking it would work there, and also macOS seems to
do it that way for as far back as I can check.

> This comment is insightful:
> https://github.com/HowardHinnant/date/issues/269#issuecomment-353792132
> It's talking about this code:
> https://github.com/HowardHinnant/date/blob/master/src/tz.cpp#L3652

Interesting.  Not sure if we want to try all those files.  But I'll
take a look at whipping up something that checks /etc/localtime.

            regards, tom lane


pgsql-hackers by date:

Previous
From: Arseny Sher
Date:
Subject: Re: Indexes on partitioned tables and foreign partitions
Next
From: Robert Haas
Date:
Subject: Re: [HACKERS] Clock with Adaptive Replacement