beit at a performance cost.
I'm thinking of this mainly for logical slots.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
--
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to y
On 3 November 2017 at 12:41, David Rowley wrote:
> On 3 November 2017 at 03:26, Craig Ringer wrote:
>> On 2 November 2017 at 22:22, David Rowley
>> wrote:
>>> Maybe, but the new implementation is not going to do well with places
>>> where we perform lcons(). Pr
r pgsql-hackers.
For password crypto please go read the SCRAM thread and the PostgreSQL
10 release notes.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
--
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresq
red statements, any WITH HOLD cursors, temp
tables, etc you were working with. Strangeness ensues.
But we now have a session-intent stuff though. So we could possibly do
it at session level.
Backends used just for a redirect would be pretty expensive though.
--
Craig Ringer
mail.com#camsr+yf0g8_fehqyfs8gsfneer9opsmovpfnidjovgqzjzh...@mail.gmail.com
It'd be nice if you summarised any outcomes from that and addressed
it, rather than taking this as a new topic.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support,
appened with pg_upgrade,
with consuming precious toast bits, and a few other things.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
--
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make c
s of list data
structures, often wrapped with a common API.
Java's Collections, the STL, you name it.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
--
Sent via pgsql-hackers mailing list (pgsql-hackers@po
ion underneath
> once before. Maybe it's time to do that again.
I know some systems use hybrid linked array-lists, where linked list
cells are multi-element.
https://en.wikipedia.org/wiki/Unrolled_linked_list
I don't have much experience with them myself.
--
Craig Ringer
7;t immutable,
> they may be considered for inlining.
> https://www.postgresql.org/message-id/5351711493487...@web53g.yandex.ru
Yep. All theoretical though, I don't think anyone (myself included)
stumped up a patch.
--
Craig Ringer http://www.2ndQuadrant.com/
Postg
unlike some of the prior protocol tweaks
I've been interested in, it'd be client-initiated so it should be
pretty safe.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
--
Sent via pgsql-hackers mailing
ould be
MERGE CONCURRENTLY ...
and when you removed the WHEN NOT MATCHED clause it'd ERROR because
that's no longer able to be done with the same concurrency-safe
semantics?
I don't know if this would be helpful TBH, or if it would negate
Simon's compatibility goals. Just another
oint used for statement level
rollback we might still have some optimisation opportunities.
Downside is that it needs support in each client driver.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
--
Sent via pgsql-hackers
into a temp table /
tuplestore / etc.
It's not clear to me why an unbounded portal fetch, using the tcp
socket windows and buffers for flow control, isn't sufficient.
Tomas, can you explain the use case a bit more?
--
Craig Ringer http://www.2ndQuadrant.com/
Pos
etaching.
These changes would make using shm_mq persistently MUCH easier,
without imposing significant cost on existing users. And it'd make it
way simpler to build a layer on top for a 1:m 2-way comms system like
Ildus is talking about.
--
Craig Ringer http://www.
in the relevant threads, so you know that. To save
others the time, see:
* https://lwn.net/Articles/724198/
* https://lwn.net/Articles/671649/
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
--
Sent via pgsql-hack
executions. Instead, the multiple result set
emulation requires the caller to 'getObject' the 'refcursor' field's
result-object, then cast it to ResultSet, and treat it as a new
(nested) result set.
True multiple result sets would be exposed in PgJDBC via getMoreResul
fetch all, and let the socket
buffering take care of things, reading results only when it wants
them, and letting the server block when the windows are full.
That's not to say that SQL-level cursor support wouldn't be nice. I'm
just trying to better understand what it's solving
On 28 October 2017 at 06:09, Michael Paquier wrote:
> On Fri, Oct 27, 2017 at 1:04 AM, Andrey Borodin wrote:
>> I'm working on backups from replication salve in WAL-G [0]
>> Backups used to use result of pg_walfile_name(pg_start_backup(...)). Call to
>> pg_start_backup() works nice, but "pg_walf
n this patch.
Hopefully he'll be able to share it.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
--
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http
ng on
> backup methods.
>
> I would find interesting to add at the bottom of the backup_label file
> a new field called "START TIMELINE: %d" to put this information in a
> more understandable shape. Any opinions?
Strong "yes" from me.
--
Craig Ringer
On 23 October 2017 at 16:16, Craig Ringer wrote:
> On 23 October 2017 at 08:30, John Lumby wrote:
>
>> All works but not perfectly -- at COMMIT, resource_owner issues
>> relcache reference leak messages about relation scans not closed
>> and also about snapshot still
some point I'd really like to expose that in a more general way so
it can be used from background workers. Right now AFAICS most
background workers have to cope with errors with a proc_exit(1) and
getting restarted to try again. Not ideal.
--
Craig Ringer http://www.2ndQu
ut here, but the context is likely something along
> the lines of externally storing all transaction ids, and periodically asking
> Postgres if they're known-to-be-aborted-by-all-transactions -- one at a time.
I think Peter is asking "why?".
--
Craig Ringer ht
;
> No interest yet, but no objections too :-)
> I'm going to add this to next commitfest.
Superficially at least, it sounds like a good idea.
We should only need a virtual xid when we're working with foreign
tables since we don't do any local heap changes.
How's it work wi
uge cross-product by the looks, and psql ran of of RAM
buffering the result. The OOM killer fired (check 'dmesg' to confirm) and
killed psql. The server noticed psql going away, and reported the fact.
None of this is surprising. What's the problem here?
--
Craig Ringer
ions as you find in the rpm spec. That can be easier
when iterating tests and builds.
Since the patches are separate, you can skip the tarball and clone the
same tag from git instead. Then apply the rpm patches as separate
commits. That's typically what I'll do, makes it easier to keep t
On 17 October 2017 at 01:02, Joshua D. Drake wrote:
> On 10/15/2017 07:39 PM, Craig Ringer wrote:
>>
>> On 13 October 2017 at 08:50, Joshua D. Drake wrote:
>>>
>>> -Hackers,
>>>
>>> I had a long call with a firm developing front end proxy/cache/
On 13 October 2017 at 08:50, Joshua D. Drake wrote:
> -Hackers,
>
> I had a long call with a firm developing front end proxy/cache/HA for
> Postgres today. Essentially the software is a replacement for PGPool in
> entirety but also supports analytics etc... When I was asking them about
> pain poin
records which is safe to skip
> (insert/update/delete records not affecting this publication).
That sounds like a giant layering violation too.
I suggest focusing on reducing the amount of work done when reading
WAL, not trying to jump over whole ranges of WAL.
--
Craig Ringer
on, but not
server_version_num, as GUC_REPORT. So if a client wants
server_version_num it has to do another round trip to query for it.
- In pg_config, where we don't expose any --version-num only --version.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x
this operation and so can point wal
> -sender to the proper LSN without decoding huge part of WAL.
> But it seems to be not so easy to implement.
Sounds like confusing layering violations to me.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 S
ys, some more subtle than others.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
--
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers
On 9 October 2017 at 15:37, Konstantin Knizhnik
wrote:
> Thank you for explanations.
>
> On 08.10.2017 16:00, Craig Ringer wrote:
>>
>> I think it'd be helpful if you provided reproduction instructions,
>> test programs, etc, making it very clear when things
your extension load via
shared_preload_libraries, registering its latch in shmem_startup_hook
.
But ... that's an off-the-cuff guess.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
--
Sent via pgsql-hacker
you seek, but start by
studying src/backend/access/heap/rewriteheap.c . Notably
logical_end_heap_rewrite, logical_rewrite_heap_tuple,
logical_begin_heap_rewrite.
At a wild "I haven't read any of the relevant code in detail yet" stab
in the dark, pg_repack is failing to do the bookkeeping re
streaming physical replication.
Hopefully that gives you something to look into, anyway. Maybe you'll
be inspired to work on parallelized logical decoding :)
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
--
S
of modules they must first install.
Meh, I don't buy that. At worst, all we have to do is provide a script
that fetches them, from distro repos if possible, and failing that
from CPAN.
With cpanminus, that's pretty darn simple too.
--
Craig Ringer http://www.2ndQuadran
hen a user has
hundreds of plpgsql functions that all like to make temp tables.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
--
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make chan
place it because it cares
about IsBootstrapProcessingMode().
(Prompted by https://dba.stackexchange.com/q/187788/7788)
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
From 011cd6e08d6d29854db637d6beb8709615f376cb Mon Sep 17 00:00
On 6 October 2017 at 08:06, Andres Freund wrote:
> On 2017-10-06 07:59:40 +0800, Craig Ringer wrote:
>> The only thing that gets me excited about a threaded postgres is the
>> ability to have a PL/Java, PL/Mono etc that don't suck. We could do
>> some really cool things
only thing that gets me excited about a threaded postgres is the
ability to have a PL/Java, PL/Mono etc that don't suck. We could do
some really cool things that just aren't practical right now.
Not compelling to a wide audience, really.
--
Craig Ringer http://www.2nd
On 3 October 2017 at 20:47, Euler Taveira wrote:
>
> 2017-10-03 5:49 GMT-03:00 Nick Dro :
> > Can someone assists with the issue posted on StackOverflow?
> >
> > https://stackoverflow.com/questions/46540537/postgresql-9-3-creation-of-group-role-causes-permission-problems
> >
> >
> > Creation of ne
ther
> > in the process?
>
> Since this patch has been in Waiting for Author state for the duration of
> the
> commitfest without moving, I’m marking it Returned with Feedback. If
> there is
> still interest in pursuing this patch, please re-submit it to the next
> commitfest with the comments addressed.
>
Thanks. I'll revisit it next CF.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
x27;m not
opposed, I just don't really see the point. I'm not seeing where it'd come
in useful.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
id. If we
get better about that, then we might need some way to ask Pg to keep extra
clog. But for now it works well enough.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
o-ordinate.
It can't do anything else, since if it unilaterally commits or rolls back
it might later find out that the nodes on the other side of the network
partition or whatever made the opposite decision and, boom!
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
built-in transaction
resolver;
then I think it's probably not going to get far.
I could see a full DTC resolver in postgres one day, once we have things
like working in-core logical rep based multi-master with 2PC support. But
that's a looong way off.
--
Craig Ringer
t impossible, but
probably irritating and verbose. And you'd have none of the DDL required to
manage it, so you'd need SQL-function equivalents.
I suspect you'd be better off tweaking pglogical to speak the same protocol
as pg10, since the pgoutput protocol is an evolution of p
noted, we can likely use pgoutput
for that to some extent at least. I think the pressing need is json, going
by the zillion plugins out there for it.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
;s done well.
That said, I'm all in favour of a generic json output plugin that shares
infrastructure with logical replication, so people who are on inflexible
environments have a fallback option. I just don't care to write it.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
>>>
>>> https://www.2ndquadrant.com/en/resources/2ndqpostgres/
>>>
>>
>> This doesn't seem like a good way to argue.
>>
>>
> Sorry, that wasn't supposed to be negative. My point was that 2ndQuadrant
> has a distribution of Post
much care for that.
Personally I'd be more friendly toward Amazon / Google / etc wanting us to
include things for their convenience if they actually usefully contributed
to development and maintenance of Pg.
--
Craig Ringer
t some time ago but ran into some issues and time
constraints. Because of the need to support older versions I'm now
committed to an approach using direct libpq connections and function calls
instead, but it seems like a real shame to do that when the replication
protocol connection is *r
or Amazon then? But now there's a new home-invented plugin that
we should adopt, ignoring any of the existing ones. Why?
https://github.com/apigee-labs/transicator/tree/master/pgoutput
>
No README?
Why did this need to be invented, rather than using an existing plugin?
I don't
t.
>
Another one to watch out for is that elog(...) and ereport(...) invoke
CHECK_FOR_INTERRUPTS. That's given me exciting surprises before when
combined with assertion checking and various exit cleanup hooks.
--
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
On 21 September 2017 at 05:50, Thomas Munro
wrote:
> On Thu, Sep 21, 2017 at 12:59 AM, Robert Haas
> wrote:
> > On Wed, Sep 20, 2017 at 5:54 AM, Craig Ringer
> wrote:
> >> By the way, dsa.c really needs a cross-reference to shm_toc.c and vice
> >> vers
On 20 September 2017 at 17:52, Craig Ringer wrote:
> On 20 September 2017 at 16:55, Thomas Munro > wrote:
>
>> On Wed, Sep 20, 2017 at 6:14 PM, Gaddam Sai Ram
>> wrote:
>> > Thank you very much! That fixed my issue! :)
>> > I was in an assumption
g
> long lived DSA area you have nothing like that.
We need, IMO, a DSA-backed heirachical MemoryContext system.
We can't use the exact MemoryContext API as-is due to the need for far
pointers though :(
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
Hi all
Here's a little utility class I wrote for value and identifier quoting for
use in TAP tests.
Might be handy for others.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
PGValues.pm
Description: Per
On 20 September 2017 at 12:16, Craig Ringer wrote:
> The thought I had in mind upthread was to get rid of logicalrep slots
>> in favor of expanding the underlying bgworker slot with some additional
>> fields that would carry whatever extra info we need about a logicalrep
>>
On 20 September 2017 at 12:06, Amit Kapila wrote:
> On Wed, Sep 20, 2017 at 9:23 AM, Tom Lane wrote:
> > Craig Ringer writes:
> >> On 19 September 2017 at 18:04, Petr Jelinek <
> petr.jeli...@2ndquadrant.com>
> >> wrote:
> >>> If y
On 20 September 2017 at 11:53, Tom Lane wrote:
> Craig Ringer writes:
> > On 19 September 2017 at 18:04, Petr Jelinek <
> petr.jeli...@2ndquadrant.com>
> > wrote:
> >> If you are asking why they are not identified by the
> >> BackgroundWorkerHandle,
On 20 September 2017 at 06:36, David Steele wrote:
>
> I just use:
>
> $SIG{__DIE__} = sub {Carp::confess @_};
>
That's what I patched into my TestLib.pm too, until I learned of
Carp::Always.
I'd rather have Carp::Always, but it's definitely an O
d (b) the postmaster remembering bgworker
registrations across crash restart with no way to tell it not to. Maybe
Petr remembers the details?
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
help.
At some point we'll also want to be able to enumerate background workers
and get handles for existing workers. Also, let background workers recover
from errors without exiting, which means factoring a bunch of stuff out of
PostgresMain. But both of those are bigger jobs.
--
Craig Ringer
emit raw flags by default, so FROZEN would't be shown at all,
only COMMITTED|INVALID. If the bool to decode combined flags is set, then
it'll show things like FROZEN, and hide COMMITTED|INVALID. Similar for
other combos.
--
Craig Ringer http://www.2ndQuadrant.com/
On 13 September 2017 at 13:44, Vaishnavi Prabakaran <
vaishnaviprabaka...@gmail.com> wrote:
> Thanks for explaining. Will change this too in next version.
>
>
Thankyou, a lot, for picking up this patch.
--
Craig Ringer http://www.2ndQuadrant.com/
Postgre
PQbatchSyncQueue
>> > +
>> > + PQbatchSyncQueue
>> > +
>> > +
>>
>> I wonder why this isn't framed as PQbatchIssue/Send/...()? Syncing seems
>> to mostly make sense from a protocol POV.
>>
>>
> Renamed to PQ
On 14 August 2017 at 11:56, Craig Ringer wrote:
>
> I don't want to block failover slots on decoding on standby just because
> decoding on standby would be nice to have.
>
However, during discussion with Tomas Munro a point has come up that does
block failover slots as cur
the new behavior was needed for internal
> > future functions since the doc wasn't changed.
>
> FWIW, I also don't think it's ok to just change the behaviour
> unconditionally and without a replacement for existing behaviour.
Seems like it just needs a new argument nowait DEFAULT false
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
s and pass it to a connect
function. I pretty much always just put the user's original connstring in
'dbname' and set expand_dbname = true instead.
It might make sense to have any new function accept PQconninfoOption*. Or a
variant of PQconninfoParse that populates k/v arrays with 'n' extra fields
allocated and zeroed on return, I guess.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
On 28 August 2017 at 19:45, Tom Lane wrote:
> Craig Ringer writes:
> > It's a pain having to find the postmaster command line to get the port
> > pg_regress started a server on. We print the port in the pg_regress
> output,
> > why not the socket directory / ho
so's anything really. null bytes aren't
usable for all scripts, and nothing else cannot also be output in the data
its self. No easy answers there. In cases where I expect that to be an
issue I sometimes use \COPY ... TO STDOUT WITH (FORMAT CSV) though.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
ong option version if -B is added, and it is
>> auto-completion friendly.
>
>
>
This doesn't really address the original issue though, that it's far from
obvious how to easily and correctly script psql.
I guess there's always the option of a docs patch for that. *shrug*
I'll see what others have to say.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
On 28 August 2017 at 15:19, Michael Paquier
wrote:
> On Mon, Aug 28, 2017 at 4:07 PM, Craig Ringer
> wrote:
> > == starting postmaster==
> > running with PID 30235; connect with:
> > psql "host='/tmp/pg_regress-j74
st 2017 at 14:08, Michael Paquier
wrote:
> On Mon, Aug 28, 2017 at 2:28 PM, Craig Ringer
> wrote:
> > It's a pain having to find the postmaster command line to get the port
> > pg_regress started a server on. We print the port in the pg_regress
> output,
> > why
an the first two, though, as I can at
least think of some cases where you might want it.
X: skip .psqlrc
Reliable, portable scripted psql shouldn't be using the local .psqlrc IMO.
It's likely to just break things in exciting ways. But I can see it being
reasonable to require this
e attached?
If you'd prefer nicer wording at the expense of two lines, maybe
running with PID 16409
connection string: 'port=50848 host=/tmp/blah'
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
From 5b242d
Hi all
I find myself regurgitating the incantation
psql -qAtX -v ON_ERRORS_STOP=1
quite a bit. It's not ... super friendly.
It strikes me that we could possibly benefit from a 'psql --batch' option.
Thoughts?
--
Craig Ringer http://www.2ndQuadrant.c
be left as stubs initially.
So the outcome would be the same, just without the assumption of specific
file name and output mechanism baked in.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
applying shm_mq to non-bgworker endpoints.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
On 21 August 2017 at 10:57, Craig Ringer wrote:
> Hi all
>
> I've noticed a possible bug / design limitation where shm_mq_wait_internal
> sleep in a latch wait forever, and the postmaster gets stuck waiting for
> the bgworker the wait is running in to exit.
>
> This ha
function pointer
called on each iteration to test whether looping should continue, to be
passed to shm_mq_attach. So if you can't supply a bgw handle, you supply
that instead. Provide a shm_mq_set_handle equivalent for it too.
Any objections to the last approach?
--
Craig Ringer
fers side by
side, or when I'm working in poor conditions where I've set my terminal to
"giant old people text" sizes, I remember the advantages of a width limit.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
tes, like the path model. With some
kind of interaction so the sub-planner for the other model could know to
generate a different sub-plan based on the context of the outer plan. I
have no idea how that could work. But I think you have about zero chance of
achieving what you want by going straight
hink it's quite a useful
> function to be used by an automated system. E.g. to ensure enough, but
> not too much, WAL is available for a tertiary standby both on the actual
> primary and a failover node.
>
I strongly agree.
If you really need to move a physical slot back (why?)
s we have macros, and I think it'd make sense
to use them here too.
Eschew direct use of HEAP_XMIN_COMMITTED, HEAP_XMIN_INVALID and
HEAP_XMIN_FROZEN in tests. Instead, consistently use HeapXminIsFrozen(),
HeapXminIsCommitted(), and HeapXminIsInvalid() or something like that.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
all xacts,
or by lower level use of the decoding code.
Reasonable?
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
erpret the bitmasks (omitting some of
> the information) assuming all the bits were set correctly.
I agree, and the patch already does half of this: it can output just the
raw bit flags, or it can interpret them to show HEAP_XMIN_FROZEN etc.
So the required change, which seems to have broad agreement, is to have the
"interpret the bits" mode show only HEAP_XMIN_FROZEN when it sees
HEAP_XMIN_COMMITTED|HEAP_XMIN_INVALID, etc. We can retain raw-flags output
as-is for when seriously bogus state is suspected.
Any takers?
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
On 15 August 2017 at 10:16, Michael Paquier
wrote:
> On Tue, Aug 15, 2017 at 11:10 AM, Craig Ringer
> wrote:
> > Ooh, this finally gives us a path toward case-insensitive default
> database
> > collation via CLDR caseLevel.
> >
> > http://userguide.i
s come first, which we see here:
>
>
Ooh, this finally gives us a path toward case-insensitive default database
collation via CLDR caseLevel.
http://userguide.icu-project.org/collation
http://www.unicode.org/reports/tr35/tr35-collation.html#Algorithm_Case
That *definitely* should be
all over
pg_stat_replication and pg_replication_slots and so on. They're already
routinely used for monitoring replication lag in bytes, waiting for a peer
to catch up, etc.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
On 15 August 2017 at 09:11, Moon Insung
wrote:
> Dear Craig Ringer
>
>
>
> Frist, thank you for implementing the necessary function.
>
>
>
> but, i have some question.
>
>
>
> question 1) vacuum freeze hint bits
>
> If run a vacuum freeze, bits in
too, just as proposed for the physical case, though no
replica->master reporting would be needed for logical failover.
So despite my initial expectations they can be moderately similar in broad
structure. But I don't think there's going to be much actual code overlap
beyond minor things like both wanting a way to query slot state on the
upstream. Both *could* use decoding on standby to advance slot positions,
but for the physical case that's just a slower (and unfinished) way to do
what we already have, wheras it's necessary for logical failover.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
t strict unit testing as the rest of
Pg's APIs aren't mocked away, but it's very practical small-unit
integration testing that helps catch issues.
I wouldn't mind having an easier and nicer way to do that built in to Pg,
but don't have many ideas about practical, low-maint
vail as
https://gist.github.com/ringerc/d4a8fe97f5fd332d8b883d596d61e257 )
To actually use the slot once decoding on standby is supported: a decoding
client on "C" can consume xacts and cause slot "X" to advance catalog_xmin,
confirmed_flush_lsn, etc. walreceiver on "C" will tell walsender on "B"
about the new slot state, and it'll get synced up-tree, then B will tell A.
Since slot is already marked permanent, state won't get copied back
down-tree, that only happens once when slot is first fully created on
master.
Some node "D" can exist as a phys rep of "C". If C fails and is replace
with D, admin can promote the down-mirror slot on "D" to an owned slot.
Make sense?
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
#x27;
and move toward higher level visibility (https://gcc.gnu.org/wiki/Visibility:
-fvisibility=hidden and __attribute__((dllexport)) ). It'd make it easier
not to forget needed PGDLLEXPORTs, let us hide stuff we consider really
internal but still share across a few files, etc.)
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
On 10 August 2017 at 23:25, Robert Haas wrote:
> On Mon, Aug 7, 2017 at 2:06 AM, Craig Ringer
> wrote:
> > I think so - specifically, that it's a leftover from a revision where the
> > xid limit was advanced before clog truncation.
> >
> > I'll be finding
On 9 August 2017 at 23:42, Robert Haas wrote:
> On Tue, Aug 8, 2017 at 4:00 AM, Craig Ringer
> wrote:
> >> - When a standby connects to a master, it can optionally supply a list
> >> of slot names that it cares about.
> >
> > Wouldn't that immediately e
; > + }
> > + if (now <= TimestampTzPlusMilliseconds(last_reply_timestamp,
> wal_sender_timeout / 2))
> > + return;
> > + }
> >
> > If not, what problem prevents?
>
> We should do CHECK_FOR_INTERRUPTS() independently of
1 - 100 of 2076 matches
Mail list logo
