pressure on shared_buffers
space, cached pages for persistent tables all sessions are sharing;
* the temp table data in shared_buffers may put pressure on shared_buffers
space for dirty buffers, forcing writes of persistent tables out earlier
therefore reducing write-combining opportunities;
-
I > SSPI > SuperStrongAuth, it has to fall back to a disconnect and
retry model like now.
* Protocols that announce supported auth methods before any kind of trust
is established make life easier for vulnerability scanners and worms
and I'm sure there are more when it comes to auth handshakes.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
ough to have a useful opinion here.
> I think that it not possible to assume that temporary data will aways fir
>> in memory.
>> So 1) seems to be not acceptable solution.
>>
>
It'd only be the metadata, but if it includes things like column histograms
and most frequent value data that'd still be undesirable to have pinned in
backend memory.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
.
> Also I have attached updated version of the global temp tables with shared
> buffers - global_shared_temp-1.patch
> It is certainly larger (~2k lines vs. 1.5k lines) because it is changing
> BufferTag and related functions.
> But I do not think that this different is so crit
On Tue, 13 Aug 2019 at 16:19, Konstantin Knizhnik
wrote:
>
>
> On 13.08.2019 8:34, Craig Ringer wrote:
>
> On Tue, 13 Aug 2019 at 00:47, Pavel Stehule
> wrote:
>
>
>> But Postgres is not storing this information now anywhere else except
>>> statistic,
y. If its exit code is 0 it will not be restarted.
You could also choose to have the worker exit with code 0 on SIGTERM, again
causing itself to be unregistered and not restarted.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
change, which is fine, but I'd be interested
in your thoughts on the question/issue in those cases.
>
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
arm)
global temp table with shared buffers becomes faster.
Who would ever do that?
I forget or do not notice some of your questions, would you be so kind as
> to repeat them?
>
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
firstval settings. Why?
+ value[SEQ_COL_LASTVAL-1] = Int64GetDatumFast(1); /* start
sequence with 1 */
Doesn't this change the test outcome for RELPERSISTENCE_UNLOGGED?:
- else if (newrelpersistence == RELPERSISTENCE_PERMANENT)
+ else if (newrelpersistence != RELPERSISTENCE_TE
ault
chunk allocations from nextval so we can avoid dealing with these
unpleasant assumptions about increment size...
... or worse, the apps that try to "fix" this by calling nextval then
setval to jump the sequence to the value they think it should have next.
And yes, I've seen this. In production code.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
ey get stuck?"
I'd rather rename it the "stuck, hard and abandoned projects list" ;)
For example I try to keep track of protocol-related stuff there, so that
if/when we do a real protocol revision we don't fail to consider things
that have come up and since been forgotten.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
ones like
check_postgres.
These issues are why I've pushed repeatedly to make server_version_num
GUC_REPORT, and expose PG_VERSION_NUM in pg_config, without success. I
still think it needs doing.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
d never unlock the replication slot used by the exiting
> > > walsender.
> >
> > Ah, sorry I was wrong. WalSndErrorCleanup() didn't get called but
> > ReplicationSlotRelease() got called. I agree that cleanup function
> > gets called in ReplicationSlotRelease().
>
> This patch is currently in 'Waiting on Author' state, but looks like it
> should actually be in Needs Review (or perhaps Ready for Commmitter)..?
>
> Craig, would you agree with that and, if so, please update the status
> accordingly.
>
>
WoA looks correct, Petr suggested a tweak to how and when the cleanup is
done that I need to adopt.
I'm just about out of time before a trip, but I'll get a colleague to
update it if I don't get the chance, so we can get it in and backpatched
for this CF.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
minor rebase.
>
I'm a smidge worried about this. It seems like psql is growing a scripting
language. Do we want to go our own way with a kind of organically grown
scripting system? Or should we be looking at embedding client-side
scripting in a more structured, formal way instead?
would casts from bit varying or bytea <-> integer and other types.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
it for years, so it's pointless to have it at all". Why add protocol v3?
PostgreSQL is a stable, long term project, and I'd like to plan for the
future. I also think people are way more likely to handle things like
--with-extra-version correctly when dealing with server_version_
> replication slot dirs, you deserve a PANIC anyway, but I'm not sure.
>
I'm happy to address those comments.
The PANIC probably made sense when it was only done on startup, but not now
it's at runtime.
The rest is mainly retained from the prior code, but it's a good cha
hy data types that are a great deal more
> complicated.
>
>
For example, an input function could conceivably take a lwlock, do some
work in the heapam, or whatever.
We don't have much in the way of rules about what input functions can or
cannot do, so you can't assume much about their behaviour and what must /
must not be cleaned up. Nor can you just reset the state in a heavy handed
manner like (say) plpgsql does.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
d someone help me get pointed in the right direction?
>
>
Nope, that's not it at all.
Logical decoding and logical rep works fine within a db instance.
It's a problem with logical decoding setup needing to see all txns open at
the time of slot creation complete before it
own version if configure finds it. Objections?
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
quot; to fail.
Sure, I can add a hack to c.h to define _THROW as a no-op when not on glibc
and all that, assuming I get far enough with this extension to bother. But
it made me ask why we have this duplication in the first place, hence this
post.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
On 29 January 2018 at 17:07, Tom Lane wrote:
>
> Another choice would be to stick a pg_ prefix on the function name.
>
That, plus a comment, seems just fine to me.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
extension, because I can just imagine how likely a patch that adds c++11
code to the core server would be to get accepted ;)
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
On 29 January 2018 at 18:02, Michael Paquier
wrote:
> On Mon, Jan 29, 2018 at 05:46:54PM +1300, Craig Ringer wrote:
> > extern "C" {
> > #include "postgres.h"
> > }
>
> Don't you need __cplusplus as well? More or less that:
>
> #ifdef
++ compiler, but port.h
reclares
+ * inet_net_ntop. If we don't annotate it the same way as the prototype in
+ * we'll upset g++, so we must use __THROW from
. If
+ * we're not on glibc, we need to define it away.
+ */
+#ifndef __GNU_LIBRARY__
+#define __THROW
+#endif
+
/*
eally viable to pursue is if it's
restricted to bind parameters. We only log those later and more selectively
as it is, so it seems much more reasonable to say "I never want to appear in the logs".
That said, I'm not sure it can be done at the function-interface lev
e all sequences are fully synced
over too.
And if 2PC is in use, the cutover must be done during a period where there
are no prepared txns on the old master, since we don't currently decode and
send txns until COMMIT PREPARED time. So we'd lose
prepared-but-not-committed txns if we cut over while they existed.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
g12 on top of a pg11 datadir with the
--allow-upgrade flag and you're done.
But I don't think I'm any keener to do the drudgery required to implement
it than anyone else is... and I share others' concerns about the
maintenance burden imposed, impact on future catalog change
BDR and pglogical. But we can patch in
a version check easily enough.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
tion locally in the tuple by setting
> tp.t_data->t_ctid.ip_blkid to InvalidBlockNumber.
>
>
We already do something similar for UPDATEs that change the REPLICA
IDENTITY; we include the oldkey in extra WAL.
The main question is whether the required knowledge is available at a
suitabl
if there's a
DB failure/outage etc. Otherwise you can never, ever, ever upgrade,
diagnostics and maintenance are harder, etc. Don't fire-and-forget. It can
be a simple producer/consumer that writes sequentially to a collection of
buffer files or whatever.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
's about getting a change
stream *out* of PostgreSQL to drive Kafka.
pglogical3 will also have some features related to feeding changes to Kafka
by the way.
> One more fact I forgot to add.. The insert load into the database is about
> 2kb/record or about 200MB/s.
>
So you're presumably rotating partitions or something; ageing things out.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
ping up the ram to 512gb as a RAM
> disk just for consistent profiling (an old bad habit).
>
> I'm reassessing my profiling plan. I'm getting ahead of myself thinking
> the OS is not having a significant effect.
>
Definitely. Especially since PostgreSQL uses buffered I/O
Pending = false;
> ProcessConfigFile(PGC_SIGHUP);
> }
>
> which'll then, in turn, send out ParameterStatus messages for changed
> GUC_REPORT GUCs.
I was wondering a while ago - can't we just set our own proc's latch here,
so we wake up and send it earlier if we're in the idle main loop?
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
_line_prefix. But then it gets
copied by pg_basebackup or whatever you're using to clone standbys etc, so
you can easily land up with multiple instances reporting the same name.
This rather defeats the purpose.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
, commit, rollback), 2pc, walsender
activity, reorderbuffer, walreceiver, slots, global xmin/catalog_xmin
changes, writes, file flushes, buffer access, and lots more. (Pg's
existing probes on transaction start and finish are almost totally
useless as you can't tell if the txn then gets an xid allocated,
whether the commit generates an xlog record or not, etc).
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
u're worried about overheads of dtrace-style probes, you can
(with systemtap ones like we use) generate a set of semaphores as a
separate .so that you link into the final build. Then you can test for
TRACE_POSTGRESQL_FOO_BAR_ENABLED() and only do any work required to
generate input for the trac
h the query results not in one batch but in a `paged`
> way, the most convenient way is to use the portals feature of PosgreSQL
> protocol.
>
>
>
Thanks. That's a really good reason. It'd also bring libpq closer to
feature-parity with PgJDBC.
Please add it to the commitfe
you can do there. It definitely won't
fly without the docs and copy/paste reduction though.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
On Fri, 8 Nov 2019 at 06:28, Mark Dilger wrote:
>
>
> On 10/31/19 10:02 AM, Andrew Dunstan wrote:
> >
> > This small patch authored by my colleague Craig Ringer enhances
> > Testlib's command_fails_like by allowing the passing of extra keyword
> > type argum
ation, 0)
WHERE l.locktype IN ('relation', 'extend', 'page', 'tuple')
UNION ALL
SELECT *
FROM pg_identify_object(l.classid, l.objid, l.objsubid)
WHERE l.locktype NOT IN ('relation', 'extend', 'page', 'tuple')
) AS lo(lock_objtype, lock_objschema, lock_objname, lock_objidentity);
$$;
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
ht and that's exactly what I mentioned in my conclusion too.
>
>
In fact, I suspect this is PostgreSQL successfully protecting itself from
an unsafe situation.
Does the host have thin-provisioned storage? lvmthin, thin-provisioned SAN,
etc?
Is the DB on NFS?
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
d've actually made more
sense to use an xlinfo flag. Probably not worth changing now though.
Be extremely careful though. If you're hiding things from logical
replication you can get all sorts of confusing and exciting results. I very
strongly suggest you make anything like this superuser-on
rt Kerberos/GSSAPI credential
> forwarding / delegation, where a client willingly forwards to the PG
> server a set of credentials which then allow the PG server to
> authenticate as that user to another system (eg: through an FDW to
> another PG server).
>
> Of course, as long as we're talking pie-in-the-sky ideas, I would
> certainly be entirely for supporting both. ;)
>
> Thanks,
>
> Stephen
>
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
On Sun, 10 Nov 2019 at 13:42, Tom Lane wrote:
> Andres Freund writes:
> > On 2019-11-08 14:49:25 +0800, Craig Ringer wrote:
> >> I recently found the need to pretty-print the contents of pg_locks. So
> >> here's a little helper to do it, for anyone else who
o fetch the sslpassword from the connection
> >> parameters, in the same way that other settings can be fetched.
> >>
> >>
> >> This is mostly the excellent work of my colleague Craig Ringer, with a
> >> few embellishments from me.
> >>
> >>
umns for the function, etc.?
>
> Okay, I have done an extra round of review, and committed it.
Thanks very much. That'll make life easier when debugging corruptions.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
On Sun, 3 Nov 2019 at 07:22, Vik Fearing
wrote:
> On 31/10/2019 08:47, Fujii Masao wrote:
> > On Mon, Oct 28, 2019 at 1:33 PM Craig Ringer
> wrote:
> >> Hi folks
> >>
> >> I was recently surprised to notice that log_line_prefix doesn't support
> a
On Wed, 18 Jul 2018 at 12:10, Kyotaro HORIGUCHI <
horiguchi.kyot...@lab.ntt.co.jp> wrote:
> At Wed, 18 Jul 2018 11:12:06 +0800, Craig Ringer
> wrote in yqtpidgg...@mail.gmail.com>
> > On 26 February 2018 at 12:06, Tsunakawa, Takayuki <
> > tsunakawa.ta...@jp.fuji
because an error was encountered when loading required library
"C:\Program Files\PostgreSQL\11\lib\libxml.dll": libxml.dll requires
"liblz.dll" but it could not be found."
Hint: not the last one.
It'll at best complain about libxml.dll being missing, despite it being
very
hysical slot, then on pg10+
the catalog_xmin sent by the replica is stored as the catalog_xmin on the
physical slot instead.
Either way, if you have hot_standby_feedback enabled on a standby, that
feedback includes the requirements of any replication slots on the standby.
--
Craig Ringer
/git.postgresql.org/gitweb/?p=postgresql.git;a=blob;f=doc/src/sgml/config.sgml;h=f83770350eda5625179526300c652f23ff29c9fe;hb=HEAD#l3400>
>
>
If this only happens when a DB is dropped under load with force, I lean
toward just documenting it as a corner case.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
a relation path
could be a valid argument is in rewriteheap.c
in logical_end_heap_rewrite(). That calls the vfd layer's FileSync() and
assumes that any failure is a fsync() syscall failure. But FileSync() can
return failure if we fail to reopen the underlying file managed by the vfd
too, per F
to suggestions, but I want things to be tolerably clean.
If someone wants a shell command wrapper, they can load a contrib that
delegates the hook to a shell command. So we can just ship a contrib, which
acts both as test coverage for the feature, and a shell-command-support
wrapper for anyone who desires that.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
hile performing our fsync() request, failed to flush,
and complained about?
I'm not confident I want to assume that.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
tReplicateId to SQL or if that's not
satisfactory, potentially replace that mechanism with the newly added one
and emulate DoNotReplicateId for BC.
I don't want two orthogonal ways to say "don't consider this for logical
replication".
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
e utility program not a psql command. See
"help".
psql->
Wording advice would be welcome.
I'd be tempted to backpatch this, since it's one of the things I see users
confused by most often now - right up there with pg_hba.conf issues,
forgetting a semicolon in psql, e
; like how to parse out the sub-parts of the string, and the
> quoting/escaping issues that will come along with that; while from
> the user's perspective it replaces a simple and intellectually-coherent
> variable definition with an unintelligible mess.
>
+1000 from me on th
larly, it might make sense to assert that we have a valid snapshot in
the SPI, which we don't presently do for read-only SPI calls. I recall that
one biting me repeatedly when I was learning this stuff.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
pe of this patch though. So personally I'm OK with
the new PGPROC field. Visibility into Pg's activity is woefully limited and
something we need to prioritize more.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
t; execute any sql orders .
>
>
Yes, that's possible. It's not easy though and I strongly suggest you look
into existing approaches like using dblink instead.
Please start a new thread rather than following an unrelated existing one.
--
Craig Ringer http://www.2
ever be when using a
historic snapshot.
It's not interesting or relevant.
Reporting the commit timestamp of the current or last-processed xact would
likely just be fonfusing. I'd rather see that in pg_stat_replication if
we're going to show it, that way we can label it usefully.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
undaries where the extra bit would push you up a variant size.
Do others see use in this?
Yes. Very, very much yes.
I'd be quick to want to expose it to SQL too.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
lar C symbol.
> You can fall back on "grep -r" or "git grep", but lots of people use
> ctags or etags or some other C-aware indexing tool.
>
>
I strongly recommend cscope with editor integration for your preferred
editor btw.
--
Craig Ringer
KEY constraints, etc).
All such management is done in C level logic with the assistance of the
pg_depend catalog and the relationships it tracks.
> Q2.1 If they are not implemented via data constraints on meta-description
> tables, why ?
>
Same as above.
> Q2.2 Is there somewhere in the documentation a list of such
> "meta-constraints" implemented by PostgreSQL?
>
Not AFAIK.
Why?
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
on, so you can semi-conveniently query
it. (I don't think we have a way for a C callable function to supply a
dynamic resultset type at plan-time to avoid the need for this, do we?
Perhaps if we use a procedure not a function?)
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
On Fri, 20 Dec 2019 at 12:18, Tom Lane wrote:
> Craig Ringer writes:
> > My understanding from reading the above is that Simon didn't propose to
> > make aborted txns visible, only in-progress uncommitted txns.
>
> Yeah, but an "in-progress uncommitted txn"
ut right
now I think most contributors are focused elsewhere.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
n
strings that use them by non-superusers. So a superuser can define a user
mapping that uses these options, but normal users may not.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
errors from OpenSSL that we pass on for example. Tricking Pg
into reading from a fifo could be problematic too.
I should've applied that restriction from the start, the same way as
passwordless connections are restricted.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
On Fri, 17 Jan 2020 at 01:09, Alexey Kondratov
wrote:
>
> > I think we shouldn't touch the paths used by replication protocol. And
> > don't we focus on how we make a change of a replication slot from SQL
> > interface persistent? It seems to me that generaly we don't need to
> > save dirty slots
_changes()
presently advances the slot position immediately, so if the client
loses its connection before reading and flushing all the data it may
be unable to recover. And while pg_logical_slot_peek_changes() lets
the app read the data w/o advancing the slot, it has to then do a
separate pg_replication_sl
to make decisions based
on what's actually replicated.
Anyone object to exporting these?
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
eat care
is required with any operator or function that could raise an error in
a row filter. But there are just so many often non-obvious ways you
can land up with an ERROR being thrown that I think it's a bit of a
user foot-gun.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
ulous regression tests to trigger some of
these behaviours. I don't recall how many of them made it into the
final patch to core but it's worth a look in the TAP test suite.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
On Wed, 22 Jan 2020 at 16:45, Ants Aasma wrote:
> The intended use case of two phase transactions is ensuring atomic
> durability of transactions across multiple database systems.
Exactly. I was trying to find a good way to say this.
It doesn't make much sense to embed a 2PC resolver in Pg unle
ned
> XA managers.
Agreed. Or use some bespoke script that does the cleanup that you
think is appropriate for your particular environment and set of bugs.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
stem's state. And if that's not enough I also want to write a
decent bloat-checking view to include in the system views, since IMO
lock-blocking, bloat, and resource retention are real monitoring pain
points right now.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
uess it's all hot air unless I pony up a patch to show
how I think it should work.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
ded about the fact that an advance is persistent
> only at the follow-up checkpoint. And the tests are fixed to not use
> a fake LSN but instead advance to the latest LSN position produced.
>
> Any objections?
LGTM. Thankyou.
--
Craig Ringer http://www.2ndQuadrant.com/
2ndQuadrant - PostgreSQL Solutions for the Enterprise
s sure it made sense at the time, but I can't explain that decision
now, and it looks like we should treat it as a failure.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
require that they not be inlined, let's not pick a method which
> makes it ambiguous and makes us have to ask "do they really want this
> limit/offset, or did they just want to make the CTE not be inlined...?"
>
>
To satisfy Tom's understandable desire to let people write queries that
behave the same on old and new versions, can we get away with back-patching
the MATERIALIZED parser enhancement as a no-op in point releases?
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
l/README where I doubt people routinely running the tests
will see it. I think I put it there mainly to help people writing/fixing
tests.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
lling Visual Studio Debugger as the hander (I always
wondered why that didn't work!) and is generally just painful. It prevents
us from collecting data via Microsoft about crashes, should we wish to do
so. And who runs Pg on windows except as a service?!
So I'm all for just removing that.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
On 20 February 2018 at 22:18, Craig Ringer wrote:
> So I'm all for just removing that.
>
... but just to be clear, about -1000 on backpatching any such thing. At
most, a new GUC that defaults to the current behaviour. But I think it's
pretty niche really.
-
doesn't make
sense; it's too niche, and too early.
I'd be in favour of leaving WER on when we find out we're in a
noninteractive service too, but that'd be a separate patch for pg11+ only.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
OT willing to
interact with git (or at the time, svn) and the other tools many of us take
for granted.
Can more be gained with user-friendly, probably web-based, translation
tools to make translation updating more accessible?
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
mmendation for such tools?
>
>
I was last involved a long time ago, so I don't have direct experience. But
I've head people speaking of how much it can help.
A look around finds a few things of interest like
http://pootle.translatehouse.org/index.html
https://trans
default to ignore/filter-out, plugins would indicate "actually I
want to choose about this one" or "I understand table rewrites". I'd prefer
not to add another change callback.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
implementation
friendly to frontend code would be frustrating, but maybe it's not as bad
as I think. Certainly if you did such a thing, many people would thank you,
because the inability to use ereport() and elog(), PG_TRY, the List API,
etc, in FRONTEND code is a constant if minor s
might be worth looking at the current logic for CHECK expressions, since
the requirements are similar. In my opinion you could safely not bother
with allowing access to user catalog tables in the filter expressions and
limit them strictly to immutable functions and the tuple its self.
--
Craig Ringe
y that might be the case, but so often it's somebody kill -9'ing the
postmaster, then removing the postmaster.pid and starting up again without
killing the workers
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
y
writeback to do.
A quick look at CheckPointGuts and CheckPointBuffers suggests we don't
presently do that but I admit I haven't read in depth.
We force a fsync of the current WAL segment when switching fsync on, but
don't do anything else AFAICS.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
and maybe refine the fsync part to get rid of false positives due to bulk
loading) and build on from there in subsequent work.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
e
how we generate relfilenode IDs, making them totally independent of the oid
space.
Unsure how practical it is, but it'd be so nice to get rid of that trap.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
initely impossible. It seems astonishingly unlikely,
but that's not always good enough.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
o be out of
> order in your email. I suggest using git format-patch, that does all
> the necessary work for you.
>
> Yep, specially git format-patch with a -v argument, so the whole patchset
is visibly versioned and sorts in the correct order.
--
Craig Ringer
nce it's minor. Still no time for libpq pipelining :(
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
tput pattern files, for example.
Easier said than done without requiring a total change in how all expected
files are written, unfortunately. "magic tags" to open a regexp section of
expected file are all well and good, but then we can't just use diff
anymore...
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
On 3 March 2018 at 13:08, Peter Eisentraut wrote:
> On 1/22/18 21:33, Craig Ringer wrote:
> > We don't have much in the way of rules about what input functions can or
> > cannot do, so you can't assume much about their behaviour and what must
> > / must not be cle
201 - 300 of 559 matches
Mail list logo
