threads listlengths 1 2.99k |
|---|
[
{
"msg_contents": "I have postgresql 9.4.8 on my server and I've noticed always growing\nmemory when I use plpython. I've made some tests and find a few\nsituations, when memory leaks. For example, when I call this procedure\nmany times, I can see an always growing memory:\n\ncreate or replace\nfunction test() ... |
[
{
"msg_contents": "Hi,\n\nat pgcon, in [1], and various other threads and conferences we talked\nabout our executor performance needing improvements to perform well in\n!OLTP workloads, and how we can get there.\n\nI've played with various techniques, on and off over the last few\nweeks. Including trying to mak... |
[
{
"msg_contents": "I complained earlier about the brute-force way that the partial\naggregation patch deals with constructing Aggrefs for the upper stage of\naggregation. It copied-and-pasted several hundred lines of setrefs.c\nso as to inject a nonstandard rule for comparing upper and lower Aggrefs.\nThat's b... |
[
{
"msg_contents": "I noticed that the EXPLAIN code is set up so that in non-text output\nmodes, you get output like this for partial-aggregate plans:\n\n \"Node Type\": \"Aggregate\", +\n \"Strategy\": \"Plain\", +\n \"Operation\": \"Finalize\", ... |
[
{
"msg_contents": "In general, moving tuplesort.c batch memory caller tuples around\nhappens when batch memory needs to be recycled, or freed outright with\npfree().\n\nI failed to take into account that CLUSTER tuplesorts need an extra\nstep when moving caller tuples to a new location (i.e. when moving\nHeapTu... |
[
{
"msg_contents": "Bjorn Munch reported off-list that this sequence:\n\nunpack tarball, cd into it\n./configure ...\ncd src/test/regress\nmake\n\nno longer works in 9.6beta2, where it did work in previous releases.\nI have confirmed both statements. The failure looks like\n\ngcc -Wall -Wmissing-prototypes -Wpo... |
[
{
"msg_contents": "Hi,\n\nas reported by Debian's OpenSSL maintainers, PostgreSQL is failing to\nbuild against a snapshot of the upcoming 1.1.0 version. The report was\nfor 9.5.3, but I can reproduce it in HEAD as well:\n\nhttps://bugs.debian.org/cgi-bin/bugreport.cgi?bug=828510\n> OpenSSL 1.1.0 is about to rel... |
[
{
"msg_contents": "Hi,\n\nI've created a background worker and I am using Postgresql-9.4. This\nbgworker handles the job queue dynamically and goes to sleep if there is no\njob to process within the next 1 hour.\n\nNow, I want to have a mechanism to wake the bgworker up in case if someone\nadds a new job while ... |
[
{
"msg_contents": "On Sun, Jun 26, 2016 at 10:35 PM, Noah Misch <noah@leadboat.com> wrote:\n>> I looked into this and found that the costs are considered fuzzily the\n>> same, and then add_path prefers the slightly-worse path on the grounds\n>> that it is marked parallel_safe while the MinMaxAgg path is not. I... |
[
{
"msg_contents": "On Sun, Jun 26, 2016 at 10:42 PM, Noah Misch <noah@leadboat.com> wrote:\n> On Mon, Jun 13, 2016 at 05:27:13PM -0400, Robert Haas wrote:\n>> One problem that I've realized that is related to this is that the way\n>> that the consider_parallel flag is being set for upper rels is almost\n>> tota... |
[
{
"msg_contents": "On Sun, Jun 26, 2016 at 10:39 PM, Noah Misch <noah@leadboat.com> wrote:\n> On Mon, Jun 13, 2016 at 04:46:19PM -0400, Tom Lane wrote:\n>> Robert Haas <robertmhaas@gmail.com> writes:\n>> > In practice, we don't yet have the ability for\n>> > parallel-safe paths from subqueries to affect plannin... |
[
{
"msg_contents": "Hi,\n\nConsider the below testcase:\n\nCREATE TABLE tab(\n c1 INT NOT NULL,\n c2 INT NOT NULL\n);\nINSERT INTO tab VALUES (1, 2);\nINSERT INTO tab VALUES (2, 1);\nINSERT INTO tab VALUES (1, 2);\n\n\ncase 1:\n\nSELECT c.c1, c.c2 from tab C WHERE c.c2 = ANY (\n SELECT 1 FROM tab A WHERE a.... |
[
{
"msg_contents": "Hello,\nsorry for my repost from psql-novice, probably it was not a right place \nfor my question.\n\nI'm trying to understand how to work with dynamic shared memory, message \nqueues and workers.\nThe problem is I can not initialize any dsm segment -\n\n void _PG_init() {\n ...\n... |
[
{
"msg_contents": "Some debug info related to my previous post -\n\n* thread #1: tid = 0x2601e9, 0x0000000100313e5e \npostgres`ResourceOwnerEnlargeDSMs + 10, queue = 'com.apple.main-thread', \nstop reason = EXC_BAD_ACCESS (code=1, address=0x130)\n * frame #0: 0x0000000100313e5e postgres`ResourceOwnerEnlargeDS... |
[
{
"msg_contents": "Postgres 9.4.5\nStreaming replica\nhot_standby_feedback=1\n\nI received an alarm that replication was lagging many hours. When I took a\nlook at what was happening on the replica I found the recovery process was\n\"stuck\" on one WAL.\n\n 8719 8717 1 Apr30 ? 15:24:49 postgres: s... |
[
{
"msg_contents": "Hi,\n\nFYI: I'd just like to point you to following two forthcoming standard\nparts from \"ISO/IEC JTS 1/SC 32\" comittee: one on JSON, and one on\n\"Multi-Dimensional Arrays\" (SQL/MDA).\n\nThey define there some things different as already in PG. See also\nPeter Baumann's slides [1] and e.g... |
[
{
"msg_contents": "Do this:\n\nCREATE DATABASE test1;\nREVOKE CONNECT ON DATABASE test1 FROM PUBLIC;\n\nRun pg_dumpall.\n\nIn 9.5, this produces\n\nCREATE DATABASE test1 WITH TEMPLATE = template0 OWNER = peter;\nREVOKE ALL ON DATABASE test1 FROM PUBLIC;\nREVOKE ALL ON DATABASE test1 FROM peter;\nGRANT ALL ON DA... |
[
{
"msg_contents": "Hello,\n\nI got the build error below. It concerns RESERVE_WAL is not defined,\nhowever, it should not be a problem to be oversight for a long time.\n\nI tried to build the latest master branch just after the switch from\nREL9_5_STABLE and \"make clean\", however, repl_gram.c was not cleaned\... |
[
{
"msg_contents": "Hi\n\nI am writing two background workers - autoreindex and scheduler. In Both I\nneed to execute queries from top level. I had to wrote redundant code\nhttps://github.com/okbob/autoreindex/blob/master/utils.c\nautoreindex_execute_sql_command\n.Same code is in pglogical. Some statements - lik... |
[
{
"msg_contents": "Hello!\n\nGot some strange behavior of random() function:\n\npostgres=# select (select random() ) from generate_series(1,10) as i;\n random\n-------------------\n 0.831577288918197\n 0.831577288918197\n 0.831577288918197\n 0.831577288918197\n 0.831577288918197\n 0.831577288918197\... |
[
{
"msg_contents": "Hi,\n\nI am testing pgbench with more than 100 connections.\nalso set max_connection in postgresql.conf more than 100.\n\nInitially pgbench tries to scale nearby 150 but later it come down to 100\nconnections and stable there.\n\nIt this limitation of pgbench? or bug? or i am doing it wrong w... |
[
{
"msg_contents": "Hi,\n\nThe document explains that pg_replication_origin_xact_reset() doesn't have\nany argument variables. But it's been actually defined so as to have two\nargument variables with pg_lsn and timestamptz data types, as follows.\n\n=# \\df pg_replication_origin_xact_reset\n ... |
[
{
"msg_contents": "Running sqlsmith on a streaming slave (master as of f8c5855) is\ninconspicuous as long as the master is idle. As soon as I start it on\nthe master as well, the standby frequently crashes in RestoreSnapshot.\nIt doesn't seem to be specific to the queries, as they don't trigger a\ncrash when r... |
[
{
"msg_contents": "Hi,\n\nCurrently the tuple returned by INSTEAD OF triggers on DELETEs is only \nused to determine whether to pretend that the DELETE happened or not, \nwhich is often not helpful enough; for example, the actual tuple might \nhave been concurrently UPDATEd by another transaction and one or mor... |
[
{
"msg_contents": "Hello,\n\nWhile I was thinking of application binary compatibility between PostgreSQL releases, some questions arose about C language user-defined functions (UDFs) and extensions that depend on them.\n\n[Q1]\nCan the same UDF binary be used with different PostgreSQL minor releases? If it is,... |
[
{
"msg_contents": "CREATE FUNCTION pg_temp.bad() RETURNS text[] LANGUAGE plpythonu AS \n$$return []$$;\nSELECT pg_temp.bad();\n bad\n-----\n {}\n(1 row)\n\nSELECT pg_temp.bad() = '{}'::text[];\n ?column?\n----------\n f\n(1 row)\n\nErm?? Turns out this is because\n\nSELECT array_dims(pg_temp.bad()), array_d... |
[
{
"msg_contents": "The link to download PostgreSQL 9.6 Beta 2 for Windows X64\nis not working.\nThe link does download something, but it's Beta 1.\n\nhttp://www.enterprisedb.com/products-services-training/pgdownload#windows\n\nThanks,\nJean-Pierre Pelletier\n\nThe link to download PostgreSQL 9.6 Beta 2 for Wind... |
[
{
"msg_contents": "https://www.postgresql.org/docs/9.6/static/using-explain.html\n\nExisting...\n\n14.1.2 Explain Analyze\n[...]\n\"\"\"\nIn some query plans, it is possible for a subplan node to be executed more\nthan once. For example, the inner index scan will be executed once per\nouter row in the above nes... |
[
{
"msg_contents": "Updating master from f8c5855..1bdae16, sqlsmith triggers \"failed to\ngenerate plan\" errors again. Below is the smallest query logged so far.\n\nregards,\nAndreas\n\n-- ERROR: plan should not reference subplan's variable\nset force_parallel_mode = 'on';\nset max_parallel_workers_per_gather... |
[
{
"msg_contents": "Hello,\n\nFor some research purposes, I am trying to modify the existing statistics\nof some tables in the catalogs in order to change the execution plan,\nexperiment with the EXPLAIN call etc.\n\nConcretely, what I'd like to do is to create a \"fake\" table with a schema\nof my choice (that'... |
[
{
"msg_contents": "Hi,\n\nwhile implementing my own C function, I mentioned that some memory is not freed by the text_overlay function in varlena.c\n\nParticularly I mean the both substrings allocated by text_substring (according to the documentation of text_substring \"The result is always a freshly palloc'd d... |
[
{
"msg_contents": "Hello,\nre-testing our application Openbravo on 9.6beta2 i found the following query failing to run with\n\nERROR: cache lookup failed for type 0\n\nTested on 9.6beta2\nSpecifically debian package version '9.6~beta2-1.pgdg+1' from apt.postgresql.org\n\nSELECT \n SUM(C_ORDERL... |
[
{
"msg_contents": "Hi all\n\nI just noticed that the Pg docs on backups don't discuss what kind of\nsnapshots are safe for use without a pg_start_backup() and pg_stop_backup()\nthen copying the extra WAL.\n\nI'd like to remedy that. My understanding is that it's safe to use a\nfilesystem or block device level s... |
[
{
"msg_contents": "Hi,\n\nI fixed typos. and attached patch for this.\nThanks.\n\nI only changed comments only in src/backend/utils/adt/tsvector_op.c\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hack... |
[
{
"msg_contents": "Hi,\n\nCurrently we have CREATE TABLE statements in a git repository that look \nroughly like this:\n\nCREATE TABLE foo(\n -- the first field\n f1 int NOT NULL,\n -- the second field\n f2 int NOT NULL,\n...\n);\n\nBut the problem is that those comments don't obviously make it all the ... |
[
{
"msg_contents": "Hello,\n\nwe have customers complaining that to_date() accepts invalid dates, and \nreturns a different date instead. This is a known issue:\n\nhttp://sql-info.de/postgresql/notes/to_date-to_timestamp-gotchas.html\n\nOn the other hand this leads to wrong dates when loading dates into the \nda... |
[
{
"msg_contents": "Hi, hackers!\n\nI think there is some room for improving GiST inserts. Following is\nthe description of what I think is performance problem.\n\nFunction gistplacetopage in file /src/backend/access/gist/gist.c is\nresponsible for updating or appending new tuple on GiST page.\nCurrently, after ... |
[
{
"msg_contents": "Hi,\n\nHere's a patch for $SUBJECT. I'll probably work on the docs a bit more \nbefore the next CF, but I thought I'd post it anyway.\n\n\n.m\n\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailp... |
[
{
"msg_contents": "It looks to me like set_append_rel_pathlist builds a partial path for\nan appendrel only if there exist partial paths for all members of the\nappendrel. I don't understand this restriction. Why could you not\nbuild it from a regular path for each appendrel child? You'd have\nto farm out en... |
[
{
"msg_contents": "Hello, this is just an entry mail for the next CF.\n\nThe tap-test fails when the soruce directoy containing spaces. I\naccidentially hit this by a Jenkins project with the name \"test\nproject\".\n\nThe function system_log() is safe for such parameters but\nbackup() uses it in wrong way. On ... |
[
{
"msg_contents": "Hi all,\n(folks interested in that in CC)\n\nWhile looking at some facility in pgcrypto, I have noticed the stanza\ncreated by 56f4478 to prevent conflicts with OpenSSL, like that:\n+#define SHA256_Init pg_SHA256_Init\n+#define SHA256_Update pg_SHA256_Update\n\nWouldn't it be better to avoid ... |
[
{
"msg_contents": "Hello everyone,\n\nRight now we have PostgreSQL on Windows Server (main data center) and\ncluster is placed on NAS. We have emergency data center on UNIX\narchitecture. We want that emergency data center could continue work on\nPostgreSQL cluster that has been used by Windows PostgreSQL.\n\nW... |
[
{
"msg_contents": "I happened to notice a bit of an inconsistency in the way initdb\nprobes max_connections and shared_buffers.\n\nThis line in the shared_buffers test:\n\n/* Use same amount of memory, independent of BLCKSZ */\ntest_buffs = (trial_bufs[i] * 8192) / BLCKSZ;\n\nhas no equivalent in the max_connec... |
[
{
"msg_contents": "Poking at NetBSD kernel source it looks like the default ulimit -s\ndepends on the architecture and ranges from 512k to 16M. Postgres\ninsists on max_stack_depth being STACK_DEPTH_SLOP -- ie 512kB -- less\nthan the ulimit setting making it impossible to start up on\narchitectures with a defau... |
[
{
"msg_contents": "Hey,\n\nFor grins while writing up an example on RLS for a client I tried the \nfollowing:\n\npolicy=# create role insert;\nCREATE ROLE\n\npolicy=# create role select;\nERROR: syntax error at or near \"select\"\nLINE 1: create role select;\n ^\npolicy=# create role delete... |
[
{
"msg_contents": "Hi,\n\nDoes anybody know what will happen if tuple queue is full, which is\nused by the parallel leader and workers? In my understanding memory\nused for the queue is a dynamic shared memory and it could be full.\n\nBest regards,\n--\nTatsuo Ishii\nSRA OSS, Inc. Japan\nEnglish: http://www.sra... |
[
{
"msg_contents": "Hi all,\n\nAttached patch fixes the typo in jsonb.c\nPlease find it.\n\nRegards,\n\n--\nMasahiko Sawada\n\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers",
"msg_date": "We... |
[
{
"msg_contents": "Hi all,\n\nI just bumped into a couple of things in visibilitymap.c:\n- visibilitymap_clear clears all bits of a visibility map, its header\ncomment mentions the contrary\n- The header of visibilitymap.c mentions correctly \"a bit\" when\nreferring to setting them, but when clearing, it shoul... |
[
{
"msg_contents": "This keeps the indentation consistent when editing the documentation\nusing Emacs.\n\n\n\n\n-- \n- Twitter seems more influential [than blogs] in the 'gets reported in\n the mainstream press' sense at least. - Matt McLeod\n- That'd be because the content of a tweet is easier to... |
[
{
"msg_contents": "Hello pgdevs,\n\nThe very minor patch attached improves the PL/pgSQL documentation about \ntrigger functions. It moves the description common to both data change & \ndatabase event triggers out of the first section and into a common header. \nIt adds a link at the beginning of the sections to... |
[
{
"msg_contents": "The files are not useful when restoring a backup and would be \nautomatically deleted on startup anyway. Also deleted an out-of-date \ncomment in dsm.c.\n\n/ Oskari\n\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://w... |
[
{
"msg_contents": "The following bug has been logged on the website:\n\nBug reference: 14230\nLogged by: Francesco Canovai\nEmail address: francesco.canovai@2ndquadrant.it\nPostgreSQL version: 9.6beta2\nOperating system: Linux\nDescription: \n\nI'm taking a concurrent backup from a s... |
[
{
"msg_contents": "Hi,\n\nI have a question regarding the source code in file pg_proc.h (postgresql-9.4.4).\n\nAt line 1224 (copied below) why the 28th identifier is timestamp_eq?\n\nDATA(insert OID = 1152 ( timestamptz_eq PGNSP PGUID 12 1 0 0 0 f f f t t f i 2 0 16 \"1184 1184\" _null_ _null_ _null_ _null_ ... |
[
{
"msg_contents": "Hi all,\n\nI was trying writing my own parallel aggregation functions on 9.6beta2.\nDuring that, we found a bit confusing behavior with SAFE/UNSAFE options.\n\nOnce a PARALLEL UNSAFE function f1_unsafe() is wrapped by\na PARALLEL SAFE function f1_safe_unsafe(), calling f1_safe_unsafe()\nprodu... |
[
{
"msg_contents": "Hi,\n\nI found $SUBJECT while trying to test parallel queries. Is this a bug?\n\n\nIn not parallel mode, EXPLAIN ANALYZE reports the information about\nSort Method as follows.\n\n=# EXPLAIN ANALYZE SELECT * FROM pgbench_accounts ORDER BY bid;\n ... |
[
{
"msg_contents": "Hi All,\n\nAs part of my masters at TAU, I'm currently conducting some research\nregarding new persistent memory technology.\nI'm using PG for this research and would like to better understand some of\nthe performance bottlenecks.\nFor this reason I'm trying to disable the WAL completely, usi... |
[
{
"msg_contents": "Hi postgresql hackers -\n\nI would like to find some analysis (published work, blog posts) on the overheads affiliated with the guarantees provided by MVCC isolation. More specifically, assuming the current workload is CPU bound (as opposed to IO) what is the CPU overhead of generating the WA... |
[
{
"msg_contents": "Hello,\n\nWhile investigating on a bloat issue with a colleague, we found that if\na bgworker executes some queries with SPI, the statistic changes will\nnever be reported, since pgstat_report_stat() is only called in regular\nbackends.\n\nIn our case, the bgworker is the only process inserti... |
[
{
"msg_contents": "\nTake the following table as an example:\n\nCREATE TABLE grue (\n id SERIAL PRIMARY KEY,\n size VARCHAR(255)\n);\nCREATE INDEX grue_size ON grue(size);\n\nNow insert approximately eleventy jillion rows, but ensure\nthat there are only about 20 distinct values for size.\n\nSELECT DISTINCT siz... |
[
{
"msg_contents": "\\timing is great.\n\\timing is helpful.\n\\timing has made me really good at mentally estimating numbers modulo\n3600000.\n\nWouldn't it be great if we had a way of printing timing in more human\nfriendly formats?\n\nAttached is a patch that allows the following (new/interesting bits in\nbol... |
[
{
"msg_contents": "Hi all,\n\nFujii-san has reminded me of the fact that we do not show in \\df+ the\nparallel status of a function. The output of \\df+ is already very\nlarge, so I guess that any people mentally sane already use it with\nthe expanded display mode, and it may not matter adding more\ninformation... |
[
{
"msg_contents": "Hi,\n\nI noticed that the return type of IsForeignScanParallelSafe described in \nfdwhandler.sgml isn't correct; that should be bool, not Size. Please \nfind attached a small patch for that.\n\nBest regards,\nEtsuro Fujita\n\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@post... |
[
{
"msg_contents": "Hi,\n\n\nI have a question about logical decoding of Postgres.\n\n\nwhere are the entry points to logical decoding?\n\nSpecifically, we want to know whether logical decoding happens immediately\nafter commit, or whether there is a polling thread that scans the Write\nAhead Log and then dumps ... |
[
{
"msg_contents": "Hello,\n\nAttached is a patch for xlogreader.c for a more informative error message\nfor allocate_recordbuf() failure.\n\nThe patch details are:\n\n\n - Project name.: None\n\n\n\n - Uniquely identifiable file name, so we can tell difference between\n your v1 and v24.: src/backend/acce... |
[
{
"msg_contents": "While testing meta-command pgbench only scripts, I noticed that there is \nan infinite loop in threadRun, which means that other tasks such as \nreporting progress do not get a chance.\n\nThe attached patch breaks this loop by always returning at the end of a \nscript.\n\nOn \"pgbench -T 3 -P... |
[
{
"msg_contents": "I wanted to test overheads on an empty query, but pgbench does not allow \nit. I do not see why not.\n\nThe attached very minor patch allows empty queries.\n\n-- \nFabien.\n\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nht... |
[
{
"msg_contents": "Minor pgbench documentation improvements so that the description is more\nprecise:\n\n - a pgbench script may not contain SQL commands, it only needs not to be\n empty.\n\n - point out explicitely variable setting meta commands.\n\n - the formula is short enough to fit on a line.\n\n-- ... |
[
{
"msg_contents": "Hello devs,\n\nI mentionned my intention to add some features to pgbench back in March:\nhttps://www.postgresql.org/message-id/alpine.DEB.2.10.1603301618570.5677@sto\n\nThe attached patch adds an \\into meta command to store results of \npreceding SELECTs into pgbench variables, so that they ... |
[
{
"msg_contents": "Currently the latency is not computed and displayed consistently:\n\n - the computation is wrong under -t (duration is zero...)\n\n - depending on the conditions it is shown with a \":\" syntax or\n a \"=\" syntax.\n\nThe attached minor patch makes the computation & display more consiste... |
[
{
"msg_contents": "the checkpoint_warning feature was added by commit 2986aa6a668bce3cfb836\nin November 2002 when we didn't have any logging of checkpointing at\nall. I propose to remove it: surely anyone who cares about analyzing\ncheckpointing behavior nowadays is using the log_checkpoint feature\ninstead, ... |
[
{
"msg_contents": "Hackers,\n\nThis just came across my twitter feed:\n\nhttps://lists.freedesktop.org/archives/systemd-devel/2014-April/018373.html\n\ntl;dr; Systemd 212 defaults to remove all IPC (including SYSV memory) \nwhen a user \"fully\" logs out.\n\nJD\n\n\n-- \nSent via pgsql-hackers mailing list (pgs... |
[
{
"msg_contents": "When you don't specify a start segment to pg_xlogdump, you get:\n\npg_xlogdump: no start log position given in range mode.\n\n\nWhat is \"range mode\", and is there any other mode for pg_xlogdump? Should\nit not just be \"no start log position or segment given\" or something like\nthat?\n\n--... |
[
{
"msg_contents": "I looked into the problem reported in bug #14238,\nhttps://www.postgresql.org/message-id/20160708151747.1426.60150@wrigleys.postgresql.org\nThe submitter was kind enough to give me a copy of the problem data,\nand it turns out that the issue is that a few of the boxes contain\nNaN coordinates... |
[
{
"msg_contents": "Hello,\n\nI am Srinivas and have been working inside PostgreSQL (mostly in the\noptimizer module and few times in the executor module). I would like to\nchange PostgreSQL code such that it also returns the cost of the optimizer\nchosen plan to the Java program (through JDBC) as part of the re... |
[
{
"msg_contents": "Currently pl/pgsql interprets the mandatory INTO of IMPORT FOREIGN\nSCHEMA as INTO variable. I estimate this to be minor oversight in\npl/pgsql parsing with respect to the introduction of this statement.\nAssuming it's easily fixed, would a patch to fix pl/pgsql parsing be\naccepted?\n\nmerl... |
[
{
"msg_contents": "btree_gin uses SEQ_MINVALUE as a way to get the smallest int64 value.\nThis is actually wrong because the smallest int64 value is\nSEQ_MINVALUE-1, so this might be slightly broken.\n\nThe whole thing was done as a convenience when INT64_IS_BUSTED had to be\nconsidered, but I think we can get ... |
[
{
"msg_contents": "Hi,\n\nI noticed odd behavior in invalidating a cached plan with a foreign join \ndue to user mapping changes. Consider:\n\npostgres=# create extension postgres_fdw;\npostgres=# create server loopback foreign data wrapper postgres_fdw \noptions (dbname 'postgres');\npostgres=# create user ... |
[
{
"msg_contents": "The following bug has been logged on the website:\n\nBug reference: 14244\nLogged by: Thomas Berger\nEmail address: thomas.berger@1und1.de\nPostgreSQL version: 9.5.3\nOperating system: any\nDescription: \n\npg_size_pretty uses the suffix \"kB\" (kilobyte, 10^3 byte... |
[
{
"msg_contents": "Hi\n\nIs there any plan to implement \"session per thread\" or \"shared\nsessions between thread\"?\nWe have analyzed the ability to contribute pgSql to jvm bytecode compiler but with\ncurrent thread model this idea is far from optimal.(Vm can be different of course.\nBut... |
[
{
"msg_contents": "I've been having some adventures with pg_basebackup lately, and had\nsome suggestions based on those.\n\nThe --help message for pg_basebackup says:\n\n-Z, --compress=0-9 compress tar output with given compression level\n\nBut -Z0 is then rejected as 'invalid compression level \"0\"'. The... |
[
{
"msg_contents": "The following bug has been logged on the website:\n\nBug reference: 14245\nLogged by: David Kellum\nEmail address: david@gravitext.com\nPostgreSQL version: 9.6beta2\nOperating system: Linux\nDescription: \n\nI am doing some (fuzz) testing of full text queries and m... |
[
{
"msg_contents": "I've noticed that pl/pgsql functions/do commands do not behave well\nwhen the statement resolves and frees memory. To be clear:\n\nFOR i in 1..1000000\nLOOP\n INSERT INTO foo VALUES (i);\nEND LOOP;\n\n...runs just fine while\n\nBEGIN\n INSERT INTO foo VALUES (1);\n INSERT INTO foo VALUES... |
[
{
"msg_contents": "Hi all,\n\nI just bumped into $subject, a variable that is never set and never used:\n--- a/src/backend/access/transam/xlog.c\n+++ b/src/backend/access/transam/xlog.c\n@@ -631,8 +631,6 @@ typedef struct XLogCtlData\n TimeLineID replayEndTLI;\n /* timestamp of last COMMIT/ABORT record r... |
[
{
"msg_contents": "Hi all,\n\nAs of now UpdateMinRecoveryPoint() is using two arguments:\n- lsn, to check if the minimum recovery point should be updated to that\n- force, a boolean flag to decide if the update should be enforced or not.\nHowever those two arguments are correlated. If lsn is\nInvalidXlogRecPtr,... |
[
{
"msg_contents": "I just noticed that the documentation for CREATE FUNCTION still mentions\r\nthat the temporary namespace is searched for functions even though that\r\nhas been removed with commit aa27977.\r\n\r\nAttached is a patch to fix that.\r\n\r\nYours,\r\nLaurenz Albe\r\n\n\n\n-- \nSent via pgsql-hacke... |
[
{
"msg_contents": "\nHi,\n\nConsider a scenario where one adds a *valid* constraint on a inheritance\nparent which is then merged with a child table's *not valid* constraint\nduring inheritance recursion. If merged, the constraint is not checked\nfor the child data even though it may have some. Is that an ove... |
[
{
"msg_contents": "The new upsert feature is a great addition, but in some cases is not\nas usable as\nI and seems lots of others would like it to be, take an example with\ncircular references:\n\ncreate table foo (\n id serial references bar(foo_id) on delete cascade,\n i int\n);\n\ncreate table bar (\n foo... |
[
{
"msg_contents": "The following bug has been logged on the website:\n\nBug reference: 14247\nLogged by: Peter Gerber\nEmail address: pgerber@tocco.ch\nPostgreSQL version: 9.5.3\nOperating system: Linux\nDescription: \n\nComments on a databases are restored on the database named the ... |
[
{
"msg_contents": "Hi,\n\nIn the thread [1] dealing with hashjoin bug introduced in 9.5, Tom \nvoiced his dislike of dense_alloc. I kinda agree with him that \nintroducing \"local allocators\" may not be the best idea, and as \ndense_alloc was introduced by me I was playing with the idea to wrap \nthis into a r... |
[
{
"msg_contents": "There are times when it would be useful to have the application_name\nconnection parameter displayed in the process name - and thus in ps and\npg_top - in addition to the user and database name.\n\nWould there be any downside to this? If it were done, are there any\nsuggestions on how it cou... |
[
{
"msg_contents": "\nHello devs,\n\nAlthough this is probably a \"feature\", it is a weird one:\n\n $ psql\n psql (9.6beta2)\n fabien=# SELECT 1 AS one \\;\n fabien-# SELECT 2 AS two ;\n two\n -----\n 2\n (1 row)\n\nWhere is my \"1\"?\n\nAlthough \"\\;\" behavior is not documented, I would have expe... |
[
{
"msg_contents": "Hi,\n\nCurrently, if old_snapshot_threshold is enabled, vacuum is prevented\nfrom truncating tables:\nstatic bool\nshould_attempt_truncation(LVRelStats *vacrelstats)\n{\n\tBlockNumber possibly_freeable;\n\n\tpossibly_freeable = vacrelstats->rel_pages - vacrelstats->nonempty_pages;\n\tif (poss... |
[
{
"msg_contents": "Current plans are to produce a 9.6beta3 release next week (ie, wrap\nMonday Jul 18 for announcement Thursday Jul 21). Keep in mind also\nthat we expect to make routine back-branch update releases on the\npreviously-announced schedule (wrap Monday Aug 8 for announcement\nThursday Aug 11). We... |
[
{
"msg_contents": "Hi all,\n\nAttached patch fixes small typo in contrib/postgres_fdw/deparse.c\n\ns/whenver/whenever/g\n\nRegards,\n\n--\nMasahiko Sawada\n\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgs... |
[
{
"msg_contents": "While running sqlsmith tool, I saw some cache lookup failure issues\nreported,\n\nWhile investigating those issues, I found one strange reason, and I feel\nIt's a bug in pg code.\n\nQuery:\npostgres=# select * from pg_catalog.pg_indexes where indexdef is not null;\nERROR: cache lookup failed... |
[
{
"msg_contents": "Hi devs,\n\nI came across a hang issue when COPY to a FIFO file, because the FIFO is\nnot opened for read on the other end. The backtrace from master branch is\nlike:\n\n#0 0x000000332ccc6c30 in __open_nocancel () from /lib64/libc.so.6\n#1 0x000000332cc6b693 in __GI__IO_file_open () from /l... |
[
{
"msg_contents": "Currently, if you run pg_xlogdump with -f, you have to specify an end\nposition in an existing file, or if you don't it will only follow until the\nend of the current file.\n\nThat seems like an oversight - if you specify -f with no end position, it\nshould follow \"into the future\" for any ... |
[
{
"msg_contents": ">>The results above are not really fair, pgbouncer.ini was a bit different on Ubuntu host (application_name_add_host was disabled). Here are the right results with exactly the same configuration:\n>>\n>>OS PostgreSQL version TPS Avg. latency\n>>RHEL 6 9.4 44898 1.425 ms\n>>RHEL 6 9.5 26199 2.... |
[
{
"msg_contents": "Hi,\n\nI installed postgres beta 2 using the RPM provided here for Cent OS.\n\nBased on this article I was trying to set *max_parallel_degree*, but\npostgres complained that there was no such config variable. But I was able\nto set *max_parallel_workers_per_gather* and test things for paralle... |
[
{
"msg_contents": ">I don't really see anything suspicious in the profile. This looks more\n>like a kernel scheduler issue than a postgres bottleneck one. It seems\n>that somehow using nonblocking IO (started in 9.5) causes scheduling\n>issues when pgbouncer is also local.\n>\n>Could you do perf stat -ddd -a sl... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.