run in to this before and in some scripts I run a CHECKPOINT before
sending a SIGINT, which reduces the size of the CHECKPOINT, but doesn't
necessarily eliminate it all of the time. In fact, I'm not sure it happens 100%
of the time either, but today I was bit by this shutdown/startup dela
fully
(supposedly).
START WAL LOCATION: 9F/3620 (file 0001009F0036)
CHECKPOINT LOCATION: 9F/37E9D6D8
BACKUP METHOD: pg_start_backup
BACKUP FROM: master
START TIME: 2012-10-06 17:48:10 UTC
LABEL: repmgr_standby_clone_1349545601
--
Sean Chittenden
s...@chittenden.org
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
ites safe, but that slaves are
still making assumptions about full page WAL entries being the MTU during some
corner cases. My observations lead me to believe that replication works fine
with full_page_writes disabled until the master cleans up a zero-filled or
damaged page.
For the arch
arios that come from a system running
out of FDs (though seek(2)'ing consuming an FD seems odd), it's more that it's
still possible for a master DB's VACUUM to clean up a bogus or partial page
write, and have the slave crash when the WAL entry is shipped over.
Are there any
alid pages
Oct 6 17:26:44 db02 postgres[19614]: [14-2] @ 19614 0: CONTEXT: xlog redo
vacuum: rel 1663/16387/20196; blk 2944, lastBlockVacuumed 2403
Oct 6 17:26:44 db02 postgres[19608]: [12-1] @ 19608 0: LOG: startup process
(PID 19614) was terminated by signal 6: Abort trap
Oct 6 17:26:4
#x27;s already doing this
optimization and is trimming it before the VERBOSE output is produced)
and grep(1) wasn't of much use, I figured I'd ask. TIA. -sc
--
Sean Chittenden
---(end of broadcast)---
TIP 6: Have you searched our list archives?
http://archives.postgresql.org
;s structure changing.
Is there a less evil way of doing what I ended up doing for views with
more than FUNC_MAX_ARGS cols? -sc
--
Sean Chittenden
---(end of broadcast)---
TIP 2: you can get off all lists at once with the unregister command
(send "unregister YourEmailAddressHere" to [EMAIL PROTECTED])
ion. I'd use a trigger, but t1 is actually a VIEW that I'm
trying to intercept commands to with RULEs. Any guidance in handling
this case is greatly appreciated. -sc
--
Sean Chittenden
---(end of broadcast)---
TIP 6: Have you searched our list archives?
http://archives.postgresql.org
_backend_id which returns
> your current slot id (not your pid) --- that would be separate
> addition.
If by slot, you mean connection ID, then this sounds like a good
compromise/patch to me. -sc
--
Sean Chittenden
---(end of broadcast)---
TI
he same token, I'd rather
have an int1 and cast for the time being, then when a solution does
pop into existence, I'll slowly either begin removing the casts or
just stop using them in future development. In the meantime, I'll
have a formally supported int1 storage type that isn't
pretty appealing to me and would be something I'd be interested
in doing the leg work for. Changing the prompt would probably be good
from a UI perspective and adding the necessary logic so that if the
connecting user had DBA privs, it'd run in a power user mode instead
of the normal data consumer mode.
-sc
--
Sean Chittenden
---(end of broadcast)---
TIP 4: Don't 'kill -9' the postmaster
'd disable any information hiding: the default would
be to hide non-critical areas including pg_catalog, pg_toast,
template1, and template0. \set POWERUSER would also work to toggle
this.. or just have \P toggle this mode. Thoughts?
-sc
--
Sean Chittenden
---(end of br
//developer.postgresql.org/docs/postgres/datatype-character.html#DATATYPE-CHARACTER-SPECIAL-TABLE
-sc
--
Sean Chittenden
---(end of broadcast)---
TIP 6: Have you searched our list archives?
http://archives.postgresql.org
team could run with depending on who
owns the copyright to the image. -sc
--
Sean Chittenden
---(end of broadcast)---
TIP 9: the planner will ignore your desire to choose an index scan if your
joining column's datatypes do not match
>1600 columns, that is an issue, but isn't one that I think can be
easily fended off without the backend being able to adapt on the fly
to different block sizes, which seems like something that could be
done with a rewrite of some of this code when table spaces are
in
uot;
Just a scratch that I finally had to scratch and ask about...
-sc
--
Sean Chittenden
---(end of broadcast)---
TIP 9: the planner will ignore your desire to choose an index scan if your
joining column's datatypes do not match
Is "index" a stop word in the archives? It doesn't seem to return any
results when typing in "index" with anything (ex: index correlation).
FWIW, just a data point. -sc
--
Sean Chittenden
---(end of broadcast)---
TIP 8
ds to be done. The connection pooling bit isn't
that bad, but it's also the part that's the most strait forward and
the bits that I'm quite familiar with. Phase II is a bit more murky
about and I'll probably have a few questions about when I get there.
Comments? The w
kind
of code and my Stevens is hidden away in one of 23 boxes from a move
earlier this month. I only know that Apache 1.3 does this with
obviously huge success on basically every *nix so it can't be too
hard.
-sc
--
Sean Chittenden
---(end of broadcast)-
resource needs.
PostgreSQL will never be single proc, multi-threaded, and I don't
think it should be for reliability's sake. See my above post,
however, as I think I may have a better way to handle "lots of
connections" without using threads. -sc
--
Sean Chittenden
-
~10K connections without any real degradation in performance. Those
stuck with select(2), you'd be lucky to get beyond a few hundred idle.
If you're unfortunate enough to be apart of the Win32 crowd...
*smack* why are you running a high volume DB server on Win32? You
should
with
newer versions of GCC and this requires you to develop your apps using
sqlrelay's API and not libpq though it's better than nothing. :)
-sc
--
Sean Chittenden
---(end of broadcast)---
TIP 5: Have you checked our extensive FAQ?
http://www.postgresql.org/docs/faqs/FAQ.html
ng about something that's done on the database side like ORA
Listener and passes active connections back to the backend that way
it's completely transparent and applies to every libpq app regardless
of the language, application, etc.
-sc
--
Sean Chittenden
---(end
and IO with the table being MVCC backed. Just UPDATE and be done
with it.
That said, it'd be cool if there was a GLOBAL TEMP table that could be
used for this kinda stuff... not having something WAL backed on this
kind of an application would make things really scream.
-sc
--
Sean Chittenden
---(end of broadcast)---
TIP 7: don't forget to increase your free space map settings
00 syslog msgs, then it becomes a given after the
> sequence wraps):
>
> CREATE FUNCTION syslog_ins(TEXT)
> RETURNS INT
> EXTERNAL SECURITY DEFINER
> AS '
> DECLARE
> a_msg ALIAS FOR $1;
> v_id syslog.id%TYPE;
> BEGIN
> v_id := NEXTVAL(&#x
syslog.id%TYPE;
BEGIN
v_id := NEXTVAL(''syslog_id_seq''::TEXT);
UPDATE syslog SET msg = a_msg WHERE id = v_id;
RETURN v_id;
' LANGUAGE 'plpgsql';
You may want to add a time component to the table/function, but I'll
leave that as an
sers at this point, or should be
if it's not.
If you are still having performance problems with the above solutions,
please send an EXPLAIN ANALYZE to the performance@ list so we can
figure out what's going on. -sc
--
Sean Chittenden
---(end of broadcast)---
TIP 4: Don't 'kill -9' the postmaster
ct, so I may have a day or so to play with this before
> the CIO figures out I'm done. :-)
>
> I'm hoping this turns out to be a tuning issue, as I'm still very
> much of a rookie at tuning PostgreSQL.
>
> I'll see if I can work something up. Should this go
org/docs/postgres/sql-fetch.html
http://developer.postgresql.org/docs/postgres/sql-close.html
--
Sean Chittenden
---(end of broadcast)---
TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]
ture RPMs should have a dependency that requires glibc
> version >= first fixed version.
Far beit for me to disagree... was this bug in only Linux's
glibc?
-sc
--
Sean Chittenden
---(end of broadcast)---
ava
isn't) and is a joy to work with, but YMMV. -sc
--
Sean Chittenden
PGP signature
---(end of broadcast)---
TIP 5: Have you checked our extensive FAQ?
http://www.postgresql.org/users-lounge/docs/faq.html
Are single quotes treated differently than double? I tried
single, but didn't have any luck... -sc
--
Sean Chittenden
PGP signature
---(end of broadcast)---
TIP 4: Don't 'kill -9' the postmaster
r: parse error at or near "-"
foo=# update pg_shadow set usename = 'www_foo' where usename = 'www-foo';
UPDATE 1
foo=# GRANT SELECT ON maildir TO www_foo;
CHANGE
Did I miss a page someplace that said I couldn't use a '-' in
a
> Unfortunately there aren't any great java ports for FreeBSD.
Check out the linux compatibility java support
linux-jdk13
I've found it to be about 95% as fast as something running
under native linux, but I get the perk of BSDs memory management and I
can typically run 1.4 times
34 matches
Mail list logo
