Re: [PATCH] Make pg_basebackup configure and start standby [Review]

Boszormenyi Zoltan <zb@cybertec.at>

From: Boszormenyi Zoltan <zb@cybertec.at>
To: Magnus Hagander <magnus@hagander.net>
Cc: Tom Lane <tgl@sss.pgh.pa.us>, Alvaro Herrera <alvherre@2ndquadrant.com>, PostgreSQL-development <pgsql-hackers@postgresql.org>
Date: 2012-11-22T16:57:01Z
Lists: pgsql-hackers

Attachments

2012-11-22 11:06 keltezéssel, Boszormenyi Zoltan írta:
> Hi,
>
> 2012-11-21 18:39 keltezéssel, Boszormenyi Zoltan írta:
>> 2012-11-21 15:29 keltezéssel, Tom Lane írta:
>>> Boszormenyi Zoltan <zb@cybertec.at> writes:
>>>> pg_basebackup needs pg_malloc() to call disconnect_and_exit(1)
>>>> instead to quit cleanly and not leave an "unexpected EOF from client"
>>>> message in the server log. Which is a macro at the moment, but
>>>> has to be turned into a real function for the reasons below.
>>> man 2 atexit
>>
>> It doesn't really help to make it easier. E.g: pg_strdup() in psql does this:
>>
>> char *
>> pg_strdup(const char *string)
>> {
>>         char       *tmp;
>>
>>         if (!string)
>>         {
>>                 psql_error("%s: pg_strdup: cannot duplicate null pointer (internal 
>> error)\n",
>>                                 pset.progname);
>>                 exit(EXIT_FAILURE);
>>         }
>>         tmp = strdup(string);
>>         if (!tmp)
>>         {
>>                 psql_error("out of memory\n");
>>                 exit(EXIT_FAILURE);
>>         }
>>         return tmp;
>> }
>>
>> The function passed to atexit() still needs to know which string to print
>> and this needs an internal variable. Also, the actions taken on pg_malloc/pg_strdup
>> errors don't produce the exact same messages.
>>
>> This is why I suggest the attached 03-pg_malloc-unified-v2.patch.
>>
>> The 04-tarutils-v2.patch unifies the common tar functions and the string functions
>> like print_val() used by _tarCreateHeader (previously _tarWriteHeader) and
>> escape_quotes().
>
> The PQconninfo patch grew its own mail thread (this is why
> there's no 1st patch in this series now) and I reworked the order
> of the patches to make it more logical.  This made the pg_basebackup
> patch 5K smaller.

The changes here are:

02-pg_malloc-unified-v4.1.patch: fixed crashing PQfinish() at exit time
in pg_basebackup, added the atexit() initializer to pg_receivexlog

03-tarutils-strutils-v3.1.patch: rebased to avoid fuzz in patching

04-pg_basebackup-v18.1.patch: use Magnus' PQconninfo patch
and filter the PQconninfoOption array internally, also rebased

>
> Best regards,
> Zoltán Böszörményi
>
>
>


-- 
----------------------------------
Zoltán Böszörményi
Cybertec Schönig & Schönig GmbH
Gröhrmühlgasse 26
A-2700 Wiener Neustadt, Austria
Web: http://www.postgresql-support.de
      http://www.postgresql.at/

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →
  1. Make the streaming replication protocol messages architecture-independent.