mes simple SSI tests can
show a lot of false positives just because of empty tables or missing
statistics (ANALYZE).
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
s
(though I don't claim that's easy). Then a SERIALIZABLE transaction
would abort if you created a conflict cycle with some other session
that has moved your cheese and it was also running in SERIALIZABLE
isolation.
[1]
https://www.postgresql.org/message-id/flat/20170605191104.1442.24999%40w
On Thu, Aug 10, 2017 at 10:26 AM, Thomas Munro
wrote:
> eaten a total of n! member space with an average size of n/2 per
Erm, math fail, not n! but 1 + 2 + ... + n.
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To m
x27;d
need to be intermittently building even larger ones. A thundering
herd of worker processes repeatedly share-locking the same row or
something like that?
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
the outcome, or some
other serialisation scheme like table or advisory locks.
[1]
https://github.com/postgres/postgres/blob/master/src/backend/executor/README#L297
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
posted earlier
to check_pg_collations for some ideas), but the idea would be
basically the same.
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
hub.com/macdice/check_pg_collations
[4]
https://www.postgresql.org/message-id/flat/85364fde-091f-bbc0-fec2-e3ede3984...@2ndquadrant.com
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
ichard Snodgrass's excellent book (which I hear is widely read at
utility companies among others), without any special library support:
http://www.cs.arizona.edu/~rts/tdbbook.pdf
His work influenced the SQL standard which I expect/hope is inspiring
those projects. SQL:2011 has a temporal fea
On Fri, Oct 14, 2016 at 2:04 AM, Kevin Grittner wrote:
> On Wed, Oct 12, 2016 at 8:06 PM, Thomas Munro
> wrote:
>> The "higher isolation levels" probably shouldn't be treated the same way.
>>
>> I think Peter's right about REPEATABLE READ. We should
On Thu, Oct 13, 2016 at 2:32 PM, Peter Geoghegan wrote:
> On Wed, Oct 12, 2016 at 6:06 PM, Thomas Munro
> wrote:
>> But yeah, the existing code raises false positive serialization
>> failures under SERIALIZABLE, and that's visible in the isolation test
>> I poste
redicate locking code via heap_fetch etc.
[1]
https://www.postgresql.org/message-id/CAEepm%3D2kYCegxp9qMR5TM1X3oXHj16iYzLPj_go52R2R07EvnA%40mail.gmail.com
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
ded the latter to report an error, but the
former seems like an oversight.
Here's a patch that shows one way to fix it. I think it does make
sense to change this, because otherwise automatic
retry-on-serialization-failure strategies will be befuddle by this
doomed transaction. And as you and Vit
n's alternative
* path) on the basis of another tuple that is not visible to MVCC snapshot.
* Check for the need to raise a serialization failure, and do so as necessary.
*/
So it seems to be working as designed. Perhaps someone could argue
that you should make an exception for tuples inse
> Close as I can come is the source version:
>
> https://www.postgresql.org/ftp/source/v8.1.18/
Ancient Red Hat source RPMs are apparently still be available for
archeology projects though:
ftp://ftp.redhat.com/pub/redhat/linux/enterprise/5Server/en/os/SRPMS/
(That's "5Server&qu
t; time.
Apparently some filesystems change the ctime for rename and others
don't, and POSIX tolerates both.
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
cluster 9.5/main. The client library
on the other hand would not be versioned in that way: there would be
just the latest major version's libpq5[2], and that is what other
things like py-psycopg2 etc would depend on (instead of depending on a
specific client major version l
sits on top and only gets its hands on
tuples emitted by nodes below it, so if there is a LIMIT then how
could it lock anything outside the limited set of rows that are
returned?
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
dependencies
among transactions
DETAIL: Reason code: Canceled on identification as a pivot, during write.
HINT: The transaction might succeed if retried.
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your
www.postgresql.org/support/versioning/
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
http://www.postgresql.org/message-id/CAEepm=3z0eolpo5wtuwsem38kbq+gjp8xxiuljkuqpm-sw7...@mail.gmail.com
That used pg_resetxlog -x $XID $PGDATA, but needed to do several hops
stop/pg_resetxlog/start hops to get all the way around the xid clock.
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-g
ould-have-its-own-syntax-in-sql/
[2]
http://www.cloudera.com/documentation/archive/impala/2-x/2-0-x/topics/impala_joins.html
[3] https://cwiki.apache.org/confluence/display/Hive/LanguageManual+Joins
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-ge
postgres=# create table dum ();
CREATE TABLE
postgres=# select * from dum;
┌──┐
├──┤
└──┘
(0 rows)
postgres=# create unique index dum_unique on dum((dum));
CREATE INDEX
postgres=# insert into dum select;
INSERT 0 1
postgres=# select * from dum;
┌──┐
├──┤
└──┘
(1 row)
postgres=# insert into dum select;
ERROR: duplicate key value violates unique constraint "dum_unique"
DETAIL: Key ((dum.*))=(()) already exists.
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
here did you get 2045 from? I thought it was like this:
number of members = number of member segment files * 1636 * 32
number of multixacts = number of offsets segment files * 2048 * 32
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.
Is this related?
http://www.postgresql.org/message-id/cak7teys9-o4bterbs3xuk2bffnnd55u2sm9j5r2fi7v6bhj...@mail.gmail.com
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
call to
WalSndWakeup in walsender.c which sets latches (= a mechanism for
waking processes) on all walsenders, and see the WaitLatchOrSocket
calls in walsender.c which wait for that to happen.
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgres
returned, it is possible that the current synchronous
standby's WAL contains the transaction or doesn't contain the
transaction, but not for you to have taken any external action based
on the commit having returned, because it didn't. (If your primary
crashes and restarts before COMMIT r
f them will actually be a synchronous standby at a time, and
the other one will take over that role if the first one is down, so
your system won't hang but you'll still have the sync standby
guarantee.
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list
On Thu, Feb 11, 2016 at 11:05 AM, Alvaro Herrera
wrote:
> Thomas Munro wrote:
>
>> 4. You could look into whether all those multixacts with many member
>> are really expected. (Large numbers of concurrent FK checks or
>> explicit share locks on the same rows perhap
But note that the freeze operations
can be IO intensive and take a long time depending on your database
size so you don't want them too often.
3. You could do nothing and wait for autovacuum to detect that you
are using more than half the member address space and trigger a
freeze, which wi
the goal is to get Unicode collation support, note also that
FreeBSD 11 (due some time this year) supports that in libc.
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
e server to report that it has fsync'ed the
WAL. (There is a patch being developed to change that so that you
might be able to wait for more than one in a future release).
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org
hat collation, you could teach
indxpath.c and (and I don't know what other planner machinery) to
consider that OID to be equivalent to DEFAULT_COLLATION_OID when
comparing them to consider an index path.
There was another email somewhere talking about constraint exclusion's
treatment o
On Tue, Oct 6, 2015 at 12:27 PM, Thomas Munro wrote:
> On Sun, Oct 4, 2015 at 11:47 PM, Michael Paquier
> wrote:
> > (Seems like you forgot to push the Reply-all button)
> >
> > On Sun, Oct 4, 2015 at 7:01 PM, Madovsky wrote:
> >> On 10/3/2015 3:30 PM, Mic
On Thu, Oct 8, 2015 at 12:49 PM, Thomas Munro wrote:
> On Thu, Oct 8, 2015 at 5:52 AM, Peter Geoghegan
> wrote:
> > On Wed, Oct 7, 2015 at 6:25 AM, Tom Lane wrote:
> >> The only real way out of such a situation is to REINDEX affected
> indexes.
> >> Refu
On Fri, Oct 23, 2015 at 12:33 PM, Dane Foster wrote:
> On Thu, Oct 22, 2015 at 6:54 PM, Thomas Munro
> wrote:
>>
>> On Fri, Oct 23, 2015 at 10:27 AM, Dane Foster wrote:
>> > On Thu, Oct 22, 2015 at 2:00 PM, Jim Nasby
>> > wrote:
>>
pecial case. Other constructs that have special behaviour for NULL
don't consider a composite type composed of NULLs to be NULL. For
example IS DISTINCT FROM, COALESCE, COUNT, STRICT functions.
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
to the first row returned by the query, or to nulls if the query
returned no rows."
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
On Thu, Oct 8, 2015 at 1:16 PM, Peter Geoghegan
wrote:
> On Wed, Oct 7, 2015 at 4:49 PM, Thomas Munro
> wrote:
>>> I agree that that would be almost as bad as carrying on, because there
>>> is no reason to think that the locale thing can easily be rolled back.
>&g
e
md5 checksums of the /usr/lib/locale/*/LC_COLLATE files or the version
of installed locale packages and automatically reindex things when
they change (I guess after restarting the cluster to clear any glibc
caches that might be lurking in long running backends). Or at least
tell me that's needed. Obvi
>> latency in case of let's say 100 hot standby.
>> it was an idea, a concept to let the master write and update the nodes, like
>> a queen bee ;)
>> but I'm afraid it's not possible, so maybe future version of pg will do it,
>> for now read from the master
On Wed, Jun 17, 2015 at 6:58 AM, Alvaro Herrera
wrote:
> Thomas Munro wrote:
>
>> Thanks. As mentioned elsewhere in the thread, I discovered that the
>> same problem exists for page boundaries, with a different error
>> message. I've tried the attached repro s
On Fri, Jun 5, 2015 at 1:47 PM, Thomas Munro
wrote:
> On Fri, Jun 5, 2015 at 11:47 AM, Thomas Munro
> wrote:
>> On Fri, Jun 5, 2015 at 9:29 AM, Robert Haas wrote:
>>> Here's a new version with some more fixes and improvements:
>>> [...]
>>
On Fri, Jun 5, 2015 at 11:47 AM, Thomas Munro
wrote:
> On Fri, Jun 5, 2015 at 9:29 AM, Robert Haas wrote:
>> Here's a new version with some more fixes and improvements:
>>
>> - SetOffsetVacuumLimit was failing to set MultiXactState->oldestOffset
>> when the ol
http://www.postgresql.org/message-id/CAEepm=1_KbHGbmPVmkUGE5qTP+B4efoCJYS0unGo-Mc5NV=u...@mail.gmail.com
I see the following during shutdown checkpoint:
LOG: could not truncate directory "pg_multixact/offsets": apparent wraparound
That message comes from SimpleLruTruncate.
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
f transaction 131072
DETAIL: Could not open file "pg_multixact/offsets/0002": No such file
or directory.
But, yeah, this isn't the bug we're looking for.
--
Thomas Munro
http://www.enterprisedb.com
checkpoint-page-boundary.sh
Description: Bourne shell script
checkpoint-segme
On Wed, Jun 3, 2015 at 3:42 PM, Alvaro Herrera wrote:
> Thomas Munro wrote:
>> On Tue, Jun 2, 2015 at 9:30 AM, Alvaro Herrera
>> wrote:
>> > My guess is that the file existed, and perhaps had one or more pages,
>> > but the wanted page doesn't exis
ster). I have also been considering a scenario
where multixact ID wraparound occurs during basebackup with some
ordering that causes trouble, but I don't yet see why it would break
if you replay the WAL from the backup label checkpoint (and I think
the repro would take days/weeks to run...)
-
ltixacts was copied. I haven't managed to get this to work (ie
produce a FATAL) and I'm out of time for a little while, but wanted to
share this idea in case it helps someone.
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
eaders (updating, locking or vacuuming). If you have
truncated multixacts referenced in your tuples then you have a
different form of corruption than the
pg_upgrade-tramples-on-oldestMultiXactId case we're trying to handle
gracefully here.
--
Thomas Munro
http://www.enterprisedb.com
tolerate-missing-offset-segments-wip.patch
Description: Binary data
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
(range->rangeStart <= range->rangeEnd &&
(segpage < range->rangeStart || segpage > range->rangeEnd)))
SlruDeleteSegment(ctl, filename);
--
Thomas Munro
http://www.enterprisedb.com
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
On 19 March 2013 01:00, Tom Lane wrote:
> Wasim Arif writes:
> > What is the road map for Postgres on the AIX platform? I understand that
> > the pg build farm contains an AIX 5.3 server; are there any plans to
> > upgrade to 6.1 and 7.1?
>
> The reason there's an AIX 5.3 buildfarm member is tha
On 28 March 2013 13:52, Shaun Thomas wrote:
> On 03/28/2013 07:43 AM, Gavan Schneider wrote:
>
> Personally I have ignored the money type in favour of numeric. Money
>> seemed to do too much behind the scenes for my taste, but, that's me
>> being lazy as well, I haven't spend much time trying to
Hi
I am using 9.1.6, and I've set up a partitioned table as described in the
manual, with partitions based on a timestamptz column called 'time'. The
exclusion constraints work nicely when I select ranges of times with
literal constants. But why would a WHERE clause like the following not
benefi
On 25 October 2012 19:46, Boris Epstein wrote:
> And if I want to split the storage - i.e., put databases into different
> directories - can I do that?
Take a look at the tablespace feature:
http://www.postgresql.org/docs/current/static/sql-createtablespace.html
You can move existing tables wit
sor is updatable. Locking is an
implementation matter (and the use of FOR UPDATE outside of a
, ie in a , may be
non-standard anyway). NOWAIT is not an ANSI SQL keyword, and
WAIT is a keyword reserved for future use.
Regards,
Thomas Munro
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
aven't
checked, but I would expect it to be the slowest build farm
member...
Thomas Munro
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general
Hi
I noticed that 'avg' works on 'interval', but 'stddev' and 'variance' don't:
hack=> create table test (start_time timestamptz, end_time timestamptz);
CREATE TABLE
hack=> insert into test values (now(), now() + interval '1 second');
INSERT 0 1`
hack=> insert into test values (now(), now() + int
Pawel Veselov wrote:
>Hi.
>
>If I have a lot (10k) tables, and each table has a btree index, and all the
>tables are being constantly inserted into, would all the indexes have to be
>in memory, and would effectively start fighting for space?
>
>Thank you,
> Pawel.
--
Sent via pgsql-general m
Adrian Klaver wrote:
>On 07/05/2012 07:46 AM, Stefan Schwarzer wrote:
>>> Now, when I launch a query which includes "crosstab()" as a postgres
>>> user, everything works fine. However, if I launch it as user XXX, it
>>> complaints:
>
The search path is indicated as:
s in a physical range quickly?
(I realise this is a pretty odd thing to want to do... I was
experimenting with a crackpot idea for storing some data in a known
physical order and finding the beginning of ends ranges by binary
chop, instead of using a btree.)
Thanks
Thomas Munro
--
Sent via
Hi
Has anyone done any work on IEEE 754-2008 decimal types for PostgreSQL?
I couldn't find anything, so I was thinking it might be a fun exercise
for learning about extending PostgreSQL with user defined types. My
first goal is to be able to store decimal numbers with a smaller disk
footprint th
61 matches
Mail list logo
