threads listlengths 1 2.99k |
|---|
[
{
"msg_contents": "Could someone translate this detail message to English:\n\n ereport(LOG,\n (errmsg(\"logical decoding found consistent point at %X/%X\",\n (uint32) (lsn >> 32), (uint32) lsn),\n errdetail(\"running xacts with xcnt == 0\")));\n\n(or ... |
[
{
"msg_contents": "I'm running 10.9.5 of OSX. \n\nI got the MySQL and PostgreSQL dependencies installed (I think). \n\nChecked out the git repo for mysql_fdw from\ngit://github.com/EnterpriseDB/mysql_fdw \n\nUSE_PGXS=1 make \n\nand got the error: \n\n mysql_fdw.c\nmysql_fdw.c:153:56: error: use of undeclared id... |
[
{
"msg_contents": "I noticed that the recent custom-path commit completely ignored my\nadvice about not including executor headers into planner headers or\nvice versa. On the way to fixing that, I was dismayed to discover\nthat the RLS patch has utterly bollixed all semblance of modularization\nof the headers.... |
[
{
"msg_contents": "It seems to me there are typos in the reference page for CREATE TABLE.\nPlease find attached a patch.\n\nThanks,\n\nBest regards,\nEtsuro Fujita\n\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mai... |
[
{
"msg_contents": "\nI often get annoyed because psql is a bit too aggressive when it decides \nwhether to put output through the pager, and the only way to avoid this \nis to turn the pager off (in which case your next query might dump many \nthousands of lines to the screen). I'd like a way to be able to spec... |
[
{
"msg_contents": "Hi,\n\nThere's a segfault when trying to dump global object from a running\n7.4.27 with a pg_dumpall of version 9.3.5 but also 9.2.9.\n\n$ pg_dumpall -g -h localhost -p 5474\n\n column number -1 is out of range 0..12\n Segmentation fault (core dumped)\n\nThe problem comes from the firs... |
[
{
"msg_contents": "\nCREATE TABLE test (\n foo text[],\n EXCLUDE USING gist (foo WITH &&)\n);\n\nERROR: data type text[] has no default operator class for access method \"gist\"\nHINT: You must specify an operator class for the index or define a default operator class for the data type.\n\nIt looks like e... |
[
{
"msg_contents": "Discussion of my incomplete group locking patch seems to have\nconverged around two points: (1) Everybody agrees that undetected\ndeadlocks are unacceptable. (2) Nobody agrees with my proposal to\ntreat locks held by group members as mutually non-conflicting. As was\nprobably evident from t... |
[
{
"msg_contents": "I was testing backwards compatibility of pg_dumpall just now, and was\nsomewhat astonished to notice the size of the output for the regression\ndatabase compared to what it was not too long ago:\n\n-rw-rw-r--. 1 tgl tgl 4509135 Nov 13 16:19 dumpall.83\n-rw-rw-r--. 1 tgl tgl 4514441 Nov 13 1... |
[
{
"msg_contents": "I have hit a rather odd issue with prepared queries on both pg 9.3 and 9.4 beta.\n\nI have this table (copy at http://samsaffron.com/testing.db.gz) with a\nvery odd performance profile:\n\nWhen I run the following prepared query it is running significantly\nslower than the raw counterpart:\n... |
[
{
"msg_contents": "Limit (cost=.... rows=20 width=175) (actual time=.... rows=20 loops=1)\n -> Sort (cost=.... rows=568733 width=175) (actual time=....\nrows=20 loops=1)\n Sort Method: top-N heapsort\n\nThe Sort estimate shows 568733 rows, whereas the actual rows are 20.\n\nBoth are correct, in a w... |
[
{
"msg_contents": "Hi,\r\n\r\nWhen we specify a value which exceeds valid range in \"Customized Options\" ,\r\nits behavior is different from \"Parameter Interaction via Configuration File\" behavior.\r\n\r\nIn case of \"Parameter Interaction via Configuration File\", it finish with FATAL error when it get thre... |
[
{
"msg_contents": "Hi,\r\n\r\n\"pg_ctl stop\" does't work propley, if --slot option is specified when WAL is flushed only it has switched.\r\nThese processes still continue even after the posmaster failed:pg_receivexlog, walsender and logger.\r\n\r\nHow to reproduce:\r\n1.Start PostgreSQL\r\n2.Create slot\r\n3.... |
[
{
"msg_contents": "Hello Hackers,\n\nAfter reading up through archives on the two $subj related TODO items\nI'm under impression that the patches[1,2] didn't make it mainly because\nof the risk of breaking SSL internals if we try to longjump out of the\nsignal handler in the middle of a blocking SSL read and/or... |
[
{
"msg_contents": "It looks like commit 0e99be1c removed the final real use of the struct\nInhPaths. Attached patch removes it entirely.\n\n-- \nPeter Geoghegan\n\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpr... |
[
{
"msg_contents": "Hi,\n\nI've just been looking into the TPC-H benchmark with intention to put\ntogether some scripts which can be run easily to output some run times for\neach of the 22 queries.\n\nI've not had a great deal of exposure to pgbench yet, but I had thought\nthat it might be able to help me run th... |
[
{
"msg_contents": "The \"special\" area in a BRIN page looks like this:\n\n> /* special space on all BRIN pages stores a \"type\" identifier */\n> #define\t\tBRIN_PAGETYPE_META\t\t\t0xF091\n> #define\t\tBRIN_PAGETYPE_REVMAP\t\t0xF092\n> #define\t\tBRIN_PAGETYPE_REGULAR\t\t0xF093\n>...\n> typedef struct BrinSpec... |
[
{
"msg_contents": "\nThis tiny change fixes what I think is a longstanding bug in psql. I \ncauses the first line of every cell to be counted twice, whereas it \nshould in fact be excluded from extra_lines / extra_row_output_lines. \nThe bug appears to date back to commit 43ee2282 in 2008. Changing it \nappears... |
[
{
"msg_contents": "Add --synchronous option to pg_receivexlog, for more reliable WAL writing.\n\nPreviously pg_receivexlog flushed WAL data only when WAL file was switched.\nThen 3dad73e added -F option to pg_receivexlog so that users could control\nhow frequently sync commands were issued to WAL files. It also... |
[
{
"msg_contents": "Hello there,\n\nI observe that the help text of vacuumdb for --analyze,\n--analyze-only, and --analyze-in-stages could do with\na little clarification in order to be self-documenting\nand thus improve the user experience of vacuumdb.\n\nThe problem is that the sole addition of the word \"only... |
[
{
"msg_contents": "Attached patch removes obsolete debugging #define in\npg_config_manual.h (RTDEBUG).\n\nThis was last used in commit 2a8d3d, from 2005. Apparently RTDEBUG is\na contraction of \"R-Tree Debug\".\n\n-- \nPeter Geoghegan\n\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.or... |
[
{
"msg_contents": "On Tue, Nov 18, 2014 at 4:31 AM, Heikki Linnakangas <hlinnakangas@vmware.com>\nwrote:\n> WAL insertion performance\n> =========================\n> To measure the performance of generating WAL, I ran the\n> wal-update-testsuite.sh that Amit also ran earlier. The cluster was\n> initialized with... |
[
{
"msg_contents": "BRIN added a new function, PageIndexDeleteNoCompact(), which is like \nPageIndexMultiDelete but does not remove unused line pointers. However, \nall the callers pass it just a single offset. So the callers would \nreally be more interested in efficiently squeezing out a single tuple \nfrom a ... |
[
{
"msg_contents": "When profiling replay the WAL generated by pgbench, I noticed the \nPageRepairFragmentation consumes a large fraction of the CPU time:\n\nPer \"perf report\":\n\n+ 33.44% 6.79% postmaster postgres [.] \nPageRepairFragmentation\n\nThe 33.44% figure includes all the functions... |
[
{
"msg_contents": "Hackers,\n\nSince querying pg_locks can be intrusive due to needing to lock the lock\npartitions, when I'm collecting data about locks I generally put a\nstatement_timeout on it. However, I'm noticing that this\nstatement_timeout appears to be completely ignored; I've seen this query\nrun fo... |
[
{
"msg_contents": "The open_datasync code opens the output file as a test to make sure the\nflags are accepted by the OS, and if it succeeds it then immediately opens\nthe file again with the same flags, overwriting and so leaking the\ndescriptor from the previous open.\n\n\nOn Windows MinGW, this prevents the ... |
[
{
"msg_contents": "[ moving discussion from -packagers to -hackers ]\n\nSummary for onlookers:\nBjorn Munch complained that the timestamptz regression tests added in\ncommit b2cbced9e failed on his Solaris machine; it emerged that he uses\n--with-system-tzdata and the tzdata files on his machine predate 2014h\n... |
[
{
"msg_contents": "Hi,\n\n> /*\n> * Set up rendering options, in particular, disable the pager, because\n> * nobody wants to be prompted while watching the output of 'watch'.\n> */\n> myopt.nullPrint = NULL;\n> myopt.topt.pager = 0;\n\nI found psql's \\watch command always ignores \\pset nu... |
[
{
"msg_contents": "I have observed in some places like exec_bind_message and exec_simple_query,\nThough these two function have already got snapshot but before they call function PortalStart,\ncurrent snapshot gets popped off and then they pass InvalidSnapshot as parameter\nbecause of which inside PortalStar... |
[
{
"msg_contents": "I observed an interesting (and I think buggy) behaviour today after one of\r\nour clusters crashed due to an \"out of space\" condition in the data directory.\r\n\r\nFive databases in that cluster have each one unlogged table.\r\n\r\nThe log reads as follows:\r\n\r\nPANIC could not write to ... |
[
{
"msg_contents": "To test WAL replay, I often set up a master-standby system with \nstreaming replication and run \"make installcheck\" on the master. \nHowever, the regression suite doesn't generate all WAL record types. I \nspent some time looking at the lcov report (make coverage-html), and \ncrafted new te... |
[
{
"msg_contents": "Currently it is possible to change the behaviour of a function with\r\nCREATE OR REPLACE FUNCTION even if the function is part of an index definition.\r\n\r\nI think that should be forbidden, because it is very likely to corrupt\r\nthe index. I expect the objection that this would break vali... |
[
{
"msg_contents": "configure.in was deprecated some years ago. We have a bug at Gentoo [1] to\nmove it to configure.ac.\n\nI've done so in my git-clone of the postgresql repo, and ran autoupdate\nto move away from the deprecated functions as well.\n\nI generated the configure script again, but that didn't chang... |
[
{
"msg_contents": "Attached is a revision of what I previously called btreecheck, which\nis now renamed to amcheck.\n\nThis is not 9.5 material - I already have 3 bigger patches in the\nqueue, 2 of which are large and complex and have major controversies,\nand one of which has details that need to be worked out... |
[
{
"msg_contents": "Hackers,\n\nOne patch that got deferred from 9.4 was the merger of recovery.conf and\npostgresql.conf, due to conflicts with ALTER SYSTEM SET. However, this\nis still a critical TODO, because recovery.conf is still an ongoing\nmajor management problem for PostgreSQL DBAs.\n\nSo, what happene... |
[
{
"msg_contents": "Hi all,\n\nJust a small thing I noticed while looking at pageinspect.sgml, the\nset of SQL examples related to BRIN indexes uses lower-case characters\nfor reserved keywords. This has been introduced by 7516f52.\nPatch is attached.\nRegards,\n-- \nMichael\n\n\n\n-- \nSent via pgsql-hackers ma... |
[
{
"msg_contents": "Hello,\n\nThe attached obvious patch adds T_CustomScan on case-switch of\nExplainTargetRel() that was oversight.\n\nIt looked like working, but what it did was just printing referenced\nname, instead of table name.\n\npostgres=# explain select ctid, * from t0 hoge where ctid > '(50,0)'::tid;\... |
[
{
"msg_contents": "Hi all,\n\nWhile playing with the GIN code for an upcoming patch, I noticed that\nwhen inserting a new entry in a new index, this code path is not able\nto make the difference if the index is in a build state or not.\nBasically, when entering in ginEntryInsert@gininsert.c GinBtree built\nvia ... |
[
{
"msg_contents": "execAmi.c points out that\n\n * (However, since the only present use of mark/restore is in mergejoin,\n * there is no need to support mark/restore in any plan type that is not\n * capable of generating ordered output. So the seqscan, tidscan,\n * and valuesscan support is actually useless co... |
[
{
"msg_contents": "> On Thu, Nov 20, 2014 at 7:10 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:\r\n> > I've done some preliminary cleanup on this patch, but I'm still pretty\r\n> > desperately unhappy about some aspects of it, in particular the way\r\n> > that it gets custom scan providers directly involved in setref... |
[
{
"msg_contents": "I thought it might be useful to add a few words at the top\nof 'src/test/regress/regress.c' to explain what it does and\nto help differentiate it from 'pg_regress.c' and\n'pg_regress_main.c'.\n\n\nRegards\n\nIan Barwick\n\n-- \n Ian Barwick http://www.2ndQuadrant.com/\n Po... |
[
{
"msg_contents": "Hi all,\n\nPlease find attached a simple patch adding fillfactor as storage parameter\nfor GIN indexes. The default value is the same as the one currently aka 100\nto have the pages completely packed when a GIN index is created.\n\nNote that to have this feature correctly working, the fix I s... |
[
{
"msg_contents": "Hello, PostgreSQL-development.\n\nThis week our support had two complains on a very similar situation.\nAfter several minutes of work clients got \"server closed the\nconnection unexpectedly. This probably means the server terminated\nabnormally before or while precessing the request.\"\n\nOn... |
[
{
"msg_contents": "pg_config_manual.h has been choosing gnu_printf as the PG_PRINTF_ATTRIBUTE for\nevery MinGW build. That invites a torrent of warnings on pre-gcc-4.4 MinGW\ncompilers, including the compiler on buildfarm member narwhal. I'm\nincreasingly using an affected compiler, because it builds twice as... |
[
{
"msg_contents": "Hi!\n\nOn production instance I have an installed preloaded library which affects \npostgresql's optimizer (online_analyze if somebody is interesting). It works \nwell except executing make installcheck of other modules, for example btree_gist:\n\ntest numeric ... ok\ntest no... |
[
{
"msg_contents": "I noticed the other day that psql doesn't honor ECHO_HIDDEN for \\sf. \nThis looks on a somewhat cursory examination to be the only significant \nplace that doesn't. Is there any reason for this, or should I just \nadjust it so that it uses PSQLexec like pretty much every other slash \ncomman... |
[
{
"msg_contents": "I tried to add BRIN index on very simple table, to test it for \"Waiting\nfor\" blogpost, btu I can't get it to work.\n\nI got newest git head, and ran:\n\n$ create table t (id int8);\nCREATE TABLE\n(depesz@[local]:5930) 20:56:22 [depesz]\n$ insert into t select generate_series(1,1000000);\nI... |
[
{
"msg_contents": "According to\nhttp://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=dromedary&dt=2014-11-21%2019%3A59%3A50\nthe order of cascaded drops at one point in the rowsecurity test is not\ntoo stable. This is not terribly surprising since it would depend on\nphysical row order in pg_depend. I thin... |
[
{
"msg_contents": "... I think it should.\n\nIt'll break some extensions, so we should warn people about it more\nprominently. Robert's text lower down in the release notes is fine,\nbut we should put a more prominent warning at the top with the other\nbackwards compatibility breakage.\n\n-- \nJosh Berkus\nPo... |
[
{
"msg_contents": "Hi,\n\nA common mistake is to try and qualify column references on the LHS of \nSET in UPDATE. The error message can be a bit cryptic at times, too. \nI've attached a patch to try and improve the situation a bit:\n\n=# create table foo(bar int);\nCREATE TABLE\n\n=# update foo set foo.bar = 1... |
[
{
"msg_contents": "Hi\n\nI did a week training and here are some ideas from people, who are starting\nwith Postgres.\n\n1. possibility to set server side variables simply as psql option.\nMotivation - more simple and natural changing datestyle for psql in shell\nscripts. \"--set\" is allocated now, but theoreti... |
[
{
"msg_contents": "Hi,\n\nPostgreSQL check-world regress failed with current GIT HEAD on my Kubuntu \n14.10.\n\nuname -a\nLinux vlD-kuci 3.16.0-24-generic #32-Ubuntu SMP Tue Oct 28 13:13:18 UTC \n2014 i686 athlon i686 GNU/Linux\n\ngdb -d /home/src/postgresql-devel/postgresql-git/postgresql/src -c core\n...\nL... |
[
{
"msg_contents": "Hi,\n\nI'm using PostgreSQL BDR 9.4beta2 to test BDR capability right now.\n\n$ psql --version\npsql (PostgreSQL) 9.4beta2\n\nI used database name bdrdemo for BDR then I've created tables with this DDL\n\nCREATE TABLE DEPARTMENT(\n ID SERIAL PRIMARY KEY NOT NULL,\n DEPT CHA... |
[
{
"msg_contents": "\nHackers,\n\nIt appears that replication connection doesn't support URI but only the\ntraditional conninfo string.\n\nsrc/backend/replication/libpqwalreceiver/libpqwalreceiver.c:99: in libpqrcv_connect():\n\n snprintf(conninfo_repl, sizeof(conninfo_repl),\n \"%s dbname=replica... |
[
{
"msg_contents": "What's the general opinion on having test programs somewhere other than\ncontrib/ ?\n\nWe currently have a number of subdirectories for test-only programs:\n\ntest_parser (a toy text search parser, added in 2007)\ndummy_seclabel (for SECURITY LABEL regression testing, added Sept 2010)\nworker... |
[
{
"msg_contents": "While checking how BM_PERMANENT flag is set (in buffer header), I noticed that\nhash index has it set too. Shouldn't pg_class(relpersistence) be 'u' in this\ncase? Currently it's set to 'p':\n\npostgres=# CREATE TABLE a(i int);\nCREATE TABLE\npostgres=# CREATE INDEX ON a USING HASH (i);\nWARN... |
[
{
"msg_contents": "Hello\n\nnow a functions with more than one polymorphic arguments are relative\nfragile due missing casting to most common type. Some our \"functions\" like\n\"coalesce\" can do it, so it is surprising for our users.\n\nour custom polymorphic function foo(anyelement, anyelement) working well ... |
[
{
"msg_contents": "Hi,\n\nI haven't seen this mentioned anywhere (although it may have as I haven't\nread through the entire history of it), but would others find it useful to\nhave ALTER SYSTEM support comments?\n\ne.g.\n\nALTER SYSTEM SET autovacuum_vacuum_scale_factor = 0.01\nWITH [ INLINE | HEADLINE ] COMME... |
[
{
"msg_contents": "All,\n\nI am simply breaking this out into its own thread from the discussion on\nadditional role attributes (\nhttp://www.postgresql.org/message-id/20141015052259.GG28859@tamriel.snowman.net\n).\n\nA few related threads/discussions/posts:\n\n*\nhttp://www.postgresql.org/message-id/2014101611... |
[
{
"msg_contents": "All,\n\nWhile reviewing the supporting documentation and functions for role\nattributes I noticed that there seems to be some confusion (at least for\nme) with CATUPDATE.\n\nThis was prompted by the following comment from Peter about\n'has_rolcatupdate' not having a superuser check.\n\nhttp:/... |
[
{
"msg_contents": "Hi all\n\nEspecially with the introduction of json support, but also in the past\nwith hstore and other things, I've sometimes found myself wishing I\ncould provide aliases in an anonymous row constructor, e.g.\n\n ROW(x AS something, y AS somethingelse)\n\nThe same thing can be done using... |
[
{
"msg_contents": "Hi,\n\nThe new WAL format calculates FPI vs plain record data like:\n\trec_len = XLogRecGetDataLen(record) + SizeOfXLogRecord;\n\tfpi_len = record->decoded_record->xl_tot_len - rec_len;\n\nDue to the amount of data now handled outside the main data portion ,\nthat doesn't seem correct to me. ... |
[
{
"msg_contents": "Hi,\n\nI found in pg_dump.c what I believe to be a typo.\n\n- * relationships are set up by doing ALTER INHERIT rather than\nusing\n+ * relationships are set up by doing ALTER TABLE INHERIT\nrather than using\n\nAttached fixes this if appropriate to do so.\n\nTha... |
[
{
"msg_contents": "> On Mon, Nov 24, 2014 at 6:57 AM, Kouhei Kaigai <kaigai@ak.jp.nec.com> wrote:\r\n> > Indeed, I don't think it is a good idea to start from this harder portion.\r\n> > Let's focus on just varno/varattno remapping to replace join relation\r\n> > by custom-scan, as an immediate target.\r\n> \r\... |
[
{
"msg_contents": "Hello everyone,\n\nFor creating indexes on more than one column, it is useful to know the sort\norder of each sort key. So now, if you run EXPLAIN in VERBOSE mode, you get\nthe sort order information in the order the sort keys are displayed - Lukas\n\n - This patch is meant for discussion\n -... |
[
{
"msg_contents": "In 9.3.5, if I set up archiving, create a database, pull a base\nbackup, look at the clock, drop database, stop the server, rm -rf\ndatadir, put back the backup, edit recovery.conf:\n\ncd /tmp; initdb foo\nedit postgresql.conf with archive_mode = on, archive_command,\nmax_wal_senders = 1, wal... |
[
{
"msg_contents": "While fooling around with the array_agg(anyarray) patch, I noted\nthat dblink_get_connections() returns NULL if there are no active\nconnections. It seems like an empty array might be a saner\ndefinition --- thoughts?\n\n\t\t\tregards, tom lane\n\n\n-- \nSent via pgsql-hackers mailing list (... |
[
{
"msg_contents": "Hi all,\n\nIn the stuff I work on in a daily basis there are a couple of\nextensions written in C++, compiling them with MSVC on Windows using\nslightly-different scripts available in src/tools after copying them\ndirectly in contrib/. However, the build scripts available in\nsrc/tools/msvc a... |
[
{
"msg_contents": "Using both 9.2.9 and 9_2_STABLE 9b468bcec15f1ea7433d4, I get a fairly\nreproducible startup failure.\n\nWhat I was doing is restore a database from a base backup and roll it\nforward with a recovery.conf until it completes and starts up. Then I\ntruncate an unlogged table and start repopulat... |
[
{
"msg_contents": "When packaging PostgreSQL for Postgres.app, I discovered a problem: strcoll doesn't work for multibyte encodings on OS X. As a consequence, text sorting in PostgreSQL doesn't work. The only workaround seemed to be to use a legacy encoding like latin1, which is inacceptable.\n\nI discovered th... |
[
{
"msg_contents": "The following bug has been logged on the website:\n\nBug reference: 12070\nLogged by: Ondřej Bouda\nEmail address: bouda@edookit.com\nPostgreSQL version: 9.4beta2\nOperating system: Windows 7 64bit\nDescription: \n\nThe hstore_to_json_loose(hstore) produces an inva... |
[
{
"msg_contents": "Hi all,\n\nWe are facing following issue in postgresql 9.1.3 in using arrow key in Solaris platform.\nCan you please help us to resolve it or any new release has fix for this or any workaround for this?\n\nissue: psql client generates a core when up arrow is used twice.\n============\nPlatfro... |
[
{
"msg_contents": "I had to make oracle_fdw work with PostgreSQL compiled using\n--with-ldap. The issue there is that Oracle's client library has the\ndelightful property of linking against a ldap library they bundle that\nhas symbol conflicts with OpenLDAP. At PostgreSQL startup libldap is\nloaded, so when lib... |
[
{
"msg_contents": "\nTom,\n\nFirst of all, thanks for your help on IRC last time with that CREATE\nINDEX memory consumption problem.\n\nAs has been pointed out in a stackexchange answer to my question[1], it\nis indeed the limitation of pre-9.4 versions, but the limit is imposed\non memtuples array, rather than... |
[
{
"msg_contents": "The following bug has been logged on the website:\n\nBug reference: 12071\nLogged by: Maxim Boguk\nEmail address: maxim.boguk@gmail.com\nPostgreSQL version: 9.2.9\nOperating system: Linux\nDescription: \n\nHi,\n\nOn one of my servers stat collector went crasy on id... |
[
{
"msg_contents": "Greetings,\n\n The include_realm default for GSSAPI and SSPI is currently\n 'include_realm=0', meaning that the realm is stripped off of the\n Kerberos principal (aka the 'system' username) prior to looking up the\n user in pg_authid.\n\n This is fine in a single-realm environment but ex... |
[
{
"msg_contents": "In the pgsql-performance thread at\nhttp://www.postgresql.org/message-id/flat/CAOR=d=3j1U_q-zf8+jUx1hkx8ps+N8pm=EUTqyFdJ5ov=+fawg@mail.gmail.com\nit emerged that a substantial part of the slowdown Scott saw from 8.4 to\n9.2 was an unexpected consequence of the fact that the planner now\nconsi... |
[
{
"msg_contents": "Tested on 9.4b3, 9.4rc1, 9.5devel\n\nselect * from json_to_record('\n{\"id\":1,\"val\":\"josh\",\"valry\":[\"potter\",\"chef\",\"programmer\"]}') as r(id\nint, val text, valry text[]);\n\nERROR: missing dimension value\n\nWith some experimentation, I can't find any way to convert a JSON arra... |
[
{
"msg_contents": "Attached is some anonymized DDL for a fairly complex schema from a \nPostgreSQL Experts client. Also attached is an explain query that runs \nagainst the schema. The client's problem is that in trying to run the \nexplain, Postgres simply runs out of memory. On my untuned 9.3 test rig, \n(Sci... |
[
{
"msg_contents": "Hi\n\npg_regress provides the command line option \"--dbname\",\nwhich is described in the help output thusly:\n\n --dbname=DB use database DB (default \"regression\")\n\nIt does however accept multiple comma separated names\nand will create a database for each name provided,\nbut AFAICS ... |
[
{
"msg_contents": "Hello there,\n\nI would like to improve error recovery of initdb when the\npassword file is empty. The present code declares \"Error 0\"\nas the cause of failure. It suffices to use ferrer() since\nfgets() returns NULL also at a premature EOF.\n\nIn addition, a minor case is the need of a lin... |
[
{
"msg_contents": "Hello there,\n\nthe text response of pg_isready is hard coded in English.\nThese short snippets really ought to be localized as well.\n\nBest regards,\n Mats Erik Andersson\n\n\n From eca5f0d8b72b7ae5508af2850977d999a0f2bca4 Mon Sep 17 00:00:00 2001\nFrom: Mats Erik Andersson <bsd@gisladiske... |
[
{
"msg_contents": "Can we tweak pg_rewind to take differential backups in PostgreSQL?\n\nI was wondering can we hack the pg_rewind code to print the details of the\nfile which have been modified compared to a target server.\nThe list output can then be used for taking differential backups.\n\nOr perhaps we can ... |
[
{
"msg_contents": "Surely that's not a value that we expect users to be able to edit. Is\npg_config_manual.h just abused as a place that's included everywhere?\n\n(I suggest utils/guc.h as a better place.)\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your sub... |
[
{
"msg_contents": "Another thing that came out of the discussion at\nhttp://www.postgresql.org/message-id/flat/CAOR=d=3j1U_q-zf8+jUx1hkx8ps+N8pm=EUTqyFdJ5ov=+fawg@mail.gmail.com\nwas that there was a significant amount of palloc/pfree traffic blamable\non the bms_first_member() loop in plpgsql's setup_param_lis... |
[
{
"msg_contents": "The other thing that emerged from\nhttp://www.postgresql.org/message-id/flat/CAOR=d=3j1U_q-zf8+jUx1hkx8ps+N8pm=EUTqyFdJ5ov=+fawg@mail.gmail.com\nis that for workloads like this, OverrideSearchPathMatchesCurrent is\na bottleneck: it's part of the code path needed to re-use a cached plan,\nand ... |
[
{
"msg_contents": "DDL deparsing is a feature that allows collection of DDL commands as they are\nexecuted in a server, in some flexible, complete, and fully-contained format\nthat allows manipulation, storage, and transmission. This feature has several\nuse cases; the two best known ones are DDL replication a... |
[
{
"msg_contents": "Hi all,\n\nWhile performing some archive recovery tests I found the following\nassertion failure once standby gets out of archive recovery (HEAD at\na5eb85e):\n(lldb)\n* thread #1: tid = 0x0000, 0x00007fff8faaf282\nlibsystem_kernel.dylib`__pthread_kill + 10, stop reason = signal\nSIGSTOP\n *... |
[
{
"msg_contents": "In light of the recent discussions about using ICU on OS X, I looked\ninto the Core Foundation locale functions (Core Foundation = traditional\nMac API in OS X, as opposed to the Unix/POSIX APIs).\n\nAttached is a proof of concept patch that just about works for the\nsorting aspects. (The ct... |
[
{
"msg_contents": "Hi all,\n\nWhen pg_event_trigger_dropped_objects is run in a context that is not\nthe one of an event trigger, currently the error code\nERRCODE_FEATURE_NOT_SUPPORTED is returned. Wouldn't it be better to\nhave an error to define an out-of-context instead? It seems that it\nwould be a good th... |
[
{
"msg_contents": "Hi,\n\nI've more than once seen that autovacuums on certain tables never\nsucceed because regular exclusive (or similar) lockers cause it to give\nway/up before finishing. Usually if some part of the application uses\nexplicit exclusive locks.\n\nIn general I think it's quite imortant that a... |
[
{
"msg_contents": "Hi,\n\nStarting a new thread which continues on from\nhttp://www.postgresql.org/message-id/CAApHDvoeC8YGWoahVSri-84eN2k0TnH6GPXp1K59y9juC1WWBg@mail.gmail.com\n\nTo give a brief summary for any new readers:\n\nThe attached patch allows for INNER JOINed relations to be removed from the\nplan,\n... |
[
{
"msg_contents": "Revert \"Add libpq function PQhostaddr().\"\n\nThis reverts commit 9f80f4835a55a1cbffcda5d23a617917f3286c14. The\nfunction returned the raw value of a connection parameter, a task served\nby PQconninfo(). The next commit will reimplement the psql \\conninfo\nchange that way. Back-patch to ... |
[
{
"msg_contents": "Revert \"Add libpq function PQhostaddr().\"\n\nThis reverts commit 9f80f4835a55a1cbffcda5d23a617917f3286c14. The\nfunction returned the raw value of a connection parameter, a task served\nby PQconninfo(). The next commit will reimplement the psql \\conninfo\nchange that way. Back-patch to ... |
[
{
"msg_contents": "Is there any way to determine the typemod of the source data for a cast? Perhaps a modification on get_call_expr_argtype(), though I'd hate to put that in an extension...\n\nBTW, it'd be nice if we better emphasized that the typmod passed to a cast function is for the destination... :/\n-- \n... |
[
{
"msg_contents": "I'm curious to know what the expected behavior is for an insert / update\nconflict on a primary key field.\n\nThe wiki suggested \"divergence conflicts\" would be logged and replication\nwould halt for the downstream master. There is a case where the databases\n\"diverge\" such that they are... |
[
{
"msg_contents": "This fixes a missing space here:\n\n http://www.postgresql.org/docs/devel/static/wal-reliability.html\n\n(present in 9.3 onwards).\n\nRegards\n\n\nIan Barwick\n\n-- \n Ian Barwick http://www.2ndQuadrant.com/\n PostgreSQL Development, 24x7 Support, Training & Services\n\n... |
[
{
"msg_contents": "Tom Lane wrote:\n> All of the MSVC critters are failing at \"make check\".\n\nYeah, I noticed that, thanks. As far as I can see the only way to fix\nit is to install dummy_seclabel to run the core seclabel test. That\ndoesn't seem smart; I think it'd be better to remove that part of the\nco... |
[
{
"msg_contents": "Hi all\n\nCurrently the client must know the size of a large lob/clob field, like\na 'bytea' or 'text' field, in order to send it to the server. This can\nforce the client to buffer all the data before sending it to the server.\n\nIt would be helpful if the v4 protocol permitted the client to... |
[
{
"msg_contents": "Initial message:\nhttp://www.postgresql.org/message-id/5447B3FF.2080406@sigaev.ru\n\nSecond version fixes a forgotten changes in pg_am.\n\n\n-- \nTeodor Sigaev E-mail: teodor@sigaev.ru\n WWW: http://www.sigae... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.