threads listlengths 1 2.99k |
|---|
[
{
"msg_contents": "It would perhaps be useful to add optional --old-confdir and \n--new-confdir parameters to pg_upgrade. If these parameters are absent \nthen pg_upgrade would work as it does now and assume that the config \nfiles are in the datadir.\n\nThe reason for this suggestion is that packages for Ubunt... |
[
{
"msg_contents": "Hi -\n\nI've had no problem upgrading copies our qa databases (which are\nbacked up and restored with pg_dump/pg_restore) but have run into the\nsame problem each time I try to upgrade a copy of our production\ndatabase (backed up and restored via PITR). After verifying a\nsuccessful restor... |
[
{
"msg_contents": "I ruined a 5 hour UPDATE by typoing a table name on a SELECT to verify\nthe update worked. I suppose I have no one else to blame, but it was\nreally frustrating, to say the least. I assume this has happened to\nothers as well.\n\nI only later found out about SAVEPOINT, which I immediately ran... |
[
{
"msg_contents": "Hi all,\n\nWhy are there no options in_regress to specify the directory where input and\noutput data are located?\nSuch options would bring more flexibility when running regressions without\nmake check/installcheck for an external application.\nIs there a particular reason for that? Or do you... |
[
{
"msg_contents": "Hello,\n\nPlease let me ask you some questions about RelCache/SysCache/CatCache \ndesign. I know I should post this to pgsql-general, but I decided to post \nhere because the content includes design questions.\n\n<<Background>>\nMy customer is facing a \"out of memory\" problem during a batch... |
[
{
"msg_contents": "Greetings postgres hackers!\n\nHere you can find www_fdw feature documentation:\nhttp://wiki.postgresql.org/wiki/WWW_FDW\n\nLooking forward for your feedback.\n\n-- \nAlexander Soudakov\nDeveloper Programmer\nemail: cygakoB@gmail.com\nskype: asudakov\n",
"msg_date": "Wed, 28 Sep 2011 17:3... |
[
{
"msg_contents": "inline\n\nOn Wed, Sep 28, 2011 at 7:21 PM, Andrew Dunstan <andrew@dunslane.net> wrote:\n>\n>\n> On 09/28/2011 09:32 AM, Alexander Soudakov wrote:\n>>\n>> Greetings postgres hackers!\n>>\n>> Here you can find www_fdw feature documentation:\n>> http://wiki.postgresql.org/wiki/WWW_FDW\n>>\n>> Lo... |
[
{
"msg_contents": "I noticed that the savepointLevel member of TransactionStateData struct is\ninitialized to 0 from TopTransactionStateData, and never incremented or\ndecremented afterwards.\n\nSince this is a file-local struct I think we can simply get rid of all\nusages of this without any risk. I visited al... |
[
{
"msg_contents": "Hi,\n\nCurrently, the mechanism that SLRU uses to truncate directory entries is\nhardcoded in SlruScanDirectory. It receives a cutoff page number and a\nboolean flag; segments found in the SLRU directory that are below the\ncutoff are deleted -- iff the flag is true.\n\nThis is fine for most... |
[
{
"msg_contents": "Hello,\n\nIf I'm not wrong, temporary tables stay in memory if they do not go over\ntemp_buffers limit (e.g. if temp_buffers is 2GB and the size of the table is\n300MB the table will remain in memory).\nWhat if a column is variable length (e.g. text), how does this column stay\nin-memory sinc... |
[
{
"msg_contents": "> On 09/10/2011 11:39 AM, Alexey Klyukin wrote:\n> Hi Andy,\n>\n> On Sep 7, 2011, at 6:40 AM, Andy Colson wrote:\n>\n> Hi Alexey, I was taking a quick look at this patch, and have a question for ya.\n>\n> ...\n>\n> Where did the other warnings go? Its right though, line 570 is bad. It also ... |
[
{
"msg_contents": "This patch would appear(?) to have languished:\n\nhttps://commitfest.postgresql.org/action/patch_view?id=541\n\nI'd really like to see it included. In the last comments of the\nreview, there seem to be problems in *terminate* backend, but even\njust pg_cancel_backend as non-superuser would b... |
[
{
"msg_contents": "I notice that we use '(none)' as a default for external_pid_file, while\nother default no-value settings in the file are ''. The attached patch\nchanges this.\n\n-- \n Bruce Momjian <bruce@momjian.us> http://momjian.us\n EnterpriseDB http://enterprisedb.... |
[
{
"msg_contents": "In researching pg_ctl -w/wait mode for pg_upgrade, I found that pg_ctl\n-w's handling of configuration-only directories is often incorrect. For\nexample, 'pg_ctl -w stop' checks for the postmaster.pid file to\ndetermine when the server is shut down, but there is no postmaster.pid\nfile in th... |
[
{
"msg_contents": "While investigating a client problem I just observed that pg_dump takes \na surprisingly large amount of time to dump a schema with a large number \nof views. The client's hardware is quite spiffy, and yet pg_dump is \ntaking many minutes to dump a schema with some 35,000 views. Here's a \nsi... |
[
{
"msg_contents": "Hi,\n\nConsider the following sequence of commands in a psql session:\n\npostgres=#create table public.sample(x int);\npostgres=#create schema new;\npostgres=#create table new.sample(x int);\npostgres=#set search_path=public,new;\n\npostgres=#\\dt\nSchema | Name | Type | Owner\n--------------... |
[
{
"msg_contents": "\nI have been investigating some build performance issues, and trying to \nnarrow down causes of slowness, and observed an odd effect, which was \nsuggested by a huge time difference between building from git and \nbuilding from a tarball.\n\nIf I do\n\n make -C src/port all\n\nand then w... |
[
{
"msg_contents": "During the discussion of Alexey Klyukin's rewrite of ParseConfigFile,\nconsiderable unhappiness was expressed by various people about the\ncomplexity and relative uselessness of the custom_variable_classes GUC.\nWhile working over his patch just now, I've come around to the side that\nwas say... |
[
{
"msg_contents": "I tried adding a not-null column in one step and got a collation\nerror for a different column. Adding the column in several steps\nworks:\n\n\nitd=> alter table livedata add column pricechanged timestamp not null default current_timestamp;\nERROR: no collation was derived for column \"whois... |
[
{
"msg_contents": "How can i get record by data block not by sql?\r\n I want to read and write lots of data by data blocks and write a record to a appointed data blocks,so i can form a disk-resident tree by recording the block address. But i don't know how to implement in postgresql. \r\n Is there system funct... |
[
{
"msg_contents": "Hello\n\nis there some possibility to get a processed rows from COPY statement\nfrom PL/pgSQL?\n\nI searched any ways, but there are no command tag.\n\nRegards\n\nPavel Stehule\n",
"msg_date": "Mon, 3 Oct 2011 16:34:04 +0200",
"msg_from": "Pavel Stehule <pavel.stehule@gmail.com>",
... |
[
{
"msg_contents": "\nThis is a subject that has come up recently, and I can think of a number \nof use cases for it.\n\nHowever, there are lots of wrinkles. For example, the names of objects \nappear in LOTS of places, and making sure we caught them all might be \nquite tricky. Say you have a table x that inher... |
[
{
"msg_contents": "Hello\n\nThere is not possible to get a number of processed rows when COPY is\nevaluated via SPI. Client can use a tag, but SPI doesn't use a tag.\n\nI propose a small change a ProcessUtility to return a processed rows.\n\nNote: I found a small inconsistency between SPI and Utility interface.... |
[
{
"msg_contents": "pg_upgrade doesn't work if the 'postgres' database has been dropped in \nthe old cluster:\n\n~/pgsql.master$ bin/pg_upgrade -b ~/pgsql.91stable/bin -B bin/ -d \n~/pgsql.91stable/data -D data-upgraded/\nPerforming Consistency Checks\n-----------------------------\nChecking current, bin, and da... |
[
{
"msg_contents": "I'm betting not, because I don't see any support for copying their\nvalues down to child processes in\nwrite_nondefault_variables/read_nondefault_variables.\n\n\t\t\tregards, tom lane\n",
"msg_date": "Tue, 04 Oct 2011 14:55:28 -0400",
"msg_from": "Tom Lane <tgl@sss.pgh.pa.us>",
"m... |
[
{
"msg_contents": "Hello,\n\nI am a program manager with the Ecosystem Engineering team at Microsoft. At //BUILD/<http://www.buildwindows.com/> in September, Microsoft announced the availability of the Windows 8 Developer Preview. Please use the resources available at the BUILD site and begin testing your appli... |
[
{
"msg_contents": "\nPostgresql 9.0.4 has the timezone:\n\n America/Blanc-Sablon\n\nHowever other sources seem to spell this with an underscore instead of dash:\n\n America/Blanc_Sablon\n\nIt appears that beside 'America/Blanc_Sablon', other multi-word timezones\nare spelled with underscore. For example: 'Aus... |
[
{
"msg_contents": "Hi,\n\nWhile the system is idle, we skip duplicate checkpoints for some\nreasons. But when wal_level is set to hot_standby, I found that\ncheckpoints are wrongly duplicated even while the system is idle.\nThe cause is that XLOG_RUNNING_XACTS WAL record always\nfollows CHECKPOINT one when wal_... |
[
{
"msg_contents": "Hi there,\n\nJust to let you know, I'm not a database expert by any means. \nI have configured dbt-2 with postgres and created a database with 4000 warehouses, \n150 customers etc. The database size is over 8G. I am aware that lock contention\ncan be checked with lockstat (and with pg_locks ... |
[
{
"msg_contents": "Hackers, \n\nHave been following with some interest the dialog recently about getting pl/python built within the v9 tree; most recently, I've been trying v9.1.1 (git) and python 3.2.2. Configure and make complete without error, but 'make install' does not produce any plpython.so \n\nI have si... |
[
{
"msg_contents": "When walsender calls out to do_pg_stop_backup() (during base backups),\nit is not possible to terminate the process with a SIGTERM - it\nrequires a SIGKILL. This can leave unkillable backends for example if\narchive_mode is on and archive_command is failing (or not set). A\nsimilar thing woul... |
[
{
"msg_contents": "Hi! *Waves*\n\nFirst off, it all seems to work as described:\n- regressions pass\n- domains work\n- tried various inherit options (merging constraints, alter table no\ninherit etc)\n- pg_dump/restore\n\nI didn't care for the changes to gram.y so I reworked it a bit so we\nnow pass is_only to ... |
[
{
"msg_contents": "tldr:\n\nSeems to be broken by\nhttp://git.postgresql.org/gitweb/?p=postgresql.git;a=commit;h=4da99ea4231e3d8bbf28b666748c1028e7b7d665\n:\ncommit 4da99ea4231e3d8bbf28b666748c1028e7b7d665\nAuthor: Robert Haas <rhaas@postgresql.org>\nDate: Mon Jun 27 10:27:17 2011 -0400\n\n Avoid having tw... |
[
{
"msg_contents": "Hi,\n\nMy apologies for the last grumpy message I wrote; I sent it out of\nsurprise that my patch was marked \"closed\" on the current commitfest.\nI hope I haven't discouraged you from reviewing this updated version.\n\nI think you intended to use the \"Waiting on Author\" status -- that\nle... |
[
{
"msg_contents": "Until recently, when using GNU tar for creating base backups, you'd have\nto ignore \"file changed as we read it\" and \"file removed before we read\nit\" warnings, which would require a bit of craftiness if you wanted to\nhide these messages while still seeing other warnings and errors from\... |
[
{
"msg_contents": "Hello\n\nI am sending a version with regress tests and basic documentation\n\nRegards\n\nPavel Stehule",
"msg_date": "Thu, 6 Oct 2011 13:52:07 +0200",
"msg_from": "Pavel Stehule <pavel.stehule@gmail.com>",
"msg_from_op": true,
"msg_subject": "patch: CHECK FUNCTION statement"
... |
[
{
"msg_contents": "Hello\n\na order of on result in foreign server test depends on locales.\n\nThis fix renames a 'cs' identifier to 's0' identifier\n\nRegards\n\nPavel Stehule\n",
"msg_date": "Thu, 6 Oct 2011 14:11:39 +0200",
"msg_from": "Pavel Stehule <pavel.stehule@gmail.com>",
"msg_from_op": tru... |
[
{
"msg_contents": "Initial Review for patch:\n\nhttp://archives.postgresql.org/pgsql-hackers/2011-09/msg00744.php\n\n\nSubmission review\n\nThe patch is in context diff format and applies cleanly to the git master. The patch includes an update to regression tests. The regression tests pass. The patch does no... |
[
{
"msg_contents": "My understanding is that libpq does not allow one to find if a result\nset column is nullable.\n\nIs this right?\n\n(I know how to get a table column nullability information from\npg_attribute.attnotnull, but when coding around the libpq API:\n\n * Is, OMG, ugly.\n\n * Doesn't cover the arb... |
[
{
"msg_contents": "Hi.\n\nI plan to extend file_fdw wrapper. I will add options to foreign\nserver: encoding, format, header, delimiter, dir.\n\nAnd i have some asks:\n- it's better to change name of extension or not\n- other suggestions\n\n\n-- \n------------\npasman\n",
"msg_date": "Thu, 6 Oct 2011 20:39:... |
[
{
"msg_contents": "On Fri, Oct 7, 2011 at 7:41 AM, Jeff Davis <pgsql@j-davis.com> wrote:\n\n> I'd prefer to include it in the initial patch. If the current GiST code\n> is going to be replaced, then there's not much sense reviewing/testing\n> it.\n>\n> You may need to consider unbounded and empty ranges special... |
[
{
"msg_contents": "Hi,\n\nI found that by default WAL_DEBUG macro has been defined in\n9.2dev and 9.1. I'm very surprised at this. Why does WAL_DEBUG\nneed to be defined by default? The performance overhead\nintroduced by WAL_DEBUG is really vanishingly low?\n\nWAL_DEBUG was defined in the following commit:\n53... |
[
{
"msg_contents": "Hello all,\nWhile working on an application I have been developing for about two \nyears now, I have come across a base-limitation in PostgreSQL for three \nseparate problems. I was talking with other members of the community \nrecently and they agreed that the issue has some merit so I thoug... |
[
{
"msg_contents": "Hi,\n\nHere's a patch to allow TOAST tables to be moved to a different \ntablespace. This item has been picked up from the TODO list.\nMain idea is to consider that a TOAST table can have its own tablespace.\n\nRegards,\n\n-- \nJT",
"msg_date": "Fri, 07 Oct 2011 17:10:30 +0200",
"msg_... |
[
{
"msg_contents": "Hi Hackers,\n\nI have a strange problem, or maybe it's not a strange problem but just\nsomething wrong with my understanding i have SIP router which works with\npostgresql using libpq, somewhere in the code it inserts a row in the\ndatabase and then when the insert is finished it invokes anot... |
[
{
"msg_contents": "Hello all,\n\nThis is a little patch to fix a typo in docs. In the length function\nshould be a space between \"string\" and \"bytea\".\n\nBest regards,\n-- \nDickson S. Guedes\nmail/xmpp: guedes@guedesoft.net - skype: guediz\nhttp://guedesoft.net - http://www.postgresql.org.br",
"msg_dat... |
[
{
"msg_contents": "Julien Tachoires 10/07/11 10:17 AM >>>\n \n> Here's a patch to allow TOAST tables to be moved to a different\n> tablespace. This item has been picked up from the TODO list.\n> Main idea is to consider that a TOAST table can have its own\n> tablespace.\n \nThanks for the patch. Please add th... |
[
{
"msg_contents": "I notice that several members of the buildfarm have been consistently\nshowing the same regression test failure since the index-only scans\npatch went in:\n\n*** /home/pgbuildfarm/workdir/HEAD/pgsql.27150/src/test/regress/expected/stats.out\tSat Oct 8 03:20:05 2011\n--- /home/pgbuildfarm/wor... |
[
{
"msg_contents": "I tried to review the \"Optimizing box_penalty\" patch:\n \nhttps://commitfest.postgresql.org/action/patch_view?id=600\n \nas posted here:\n \nhttp://archives.postgresql.org/message-id/4E088690.5080706@enterprisedb.com\n \nThe patch no longer applies to source code, due to other recent GiST\n... |
[
{
"msg_contents": "It seems that in order to create an object in a given schema, I must\nhave been granted create privilege on the schema. But in order to drop\nthat object I require usage privilege. \n\nThis means that with the right privilege settings I can create objects\nthat I cannot subsequently drop, o... |
[
{
"msg_contents": "Florian Pflug wrote:\n \n> Coming up with a reasonable algorithm isn't *that* hard.\n \nAgreed. Our shop has used a home-grown framework for over a decade\nwhere we parse queries using ANTLR ( http://www.antlr.org/ ) and we\ntracked this trough all expressions. There really weren't that ma... |
[
{
"msg_contents": "[ Thanks all for the very productive discussion in the thread\n \"libpq, PQdescribePrepared -> PQftype, PQfmod, no PQnullable\"\n which I originated. Very useful. Now on something different. ]\n \nAbout two years ago, I had to research some PostgreSQL failures on\nHP-UX on a lame PA-RISC... |
[
{
"msg_contents": "Hello\n\nthere is fix a locale dependency of regress tests\n\nLast time I forgot to attach a patch\n\nRegards\n\nPavel Stehule",
"msg_date": "Mon, 10 Oct 2011 08:34:26 +0200",
"msg_from": "Pavel Stehule <pavel.stehule@gmail.com>",
"msg_from_op": true,
"msg_subject": "patch: f... |
[
{
"msg_contents": "Attached patch.\n\n\n\n\n------------\npasman",
"msg_date": "Mon, 10 Oct 2011 09:08:42 +0200",
"msg_from": "=?ISO-8859-2?Q?pasman_pasma=F1ski?= <pasman.p@gmail.com>",
"msg_from_op": true,
"msg_subject": "Extend file_fdw wrapper"
},
{
"msg_contents": "2011/10/10 pasman ... |
[
{
"msg_contents": "Florian Pflug wrote:\n> On Oct9, 2011, at 14:20 , Kevin Grittner wrote:\n>> Florian Pflug wrote:\n>> \n>>> Coming up with a reasonable algorithm isn't *that* hard.\n>> \n>> Agreed. Our shop has used a home-grown framework for over a decade\n>> where we parse queries using ANTLR ( http://www.... |
[
{
"msg_contents": "OK, I'm stumped:\n\nrhaas=# create extension pg_stat_statements;\nCREATE EXTENSION\nrhaas=# drop view pg_stat_statements;\nERROR: cannot drop view pg_stat_statements because extension\npg_stat_statements requires it\nHINT: You can drop extension pg_stat_statements instead.\nrhaas=# alter ex... |
[
{
"msg_contents": "Hi.\n\nCurrent behaviour is error message when foreign table and foreign\nserver have the same option defined.\n\nI don't know how to write regression test, may i read about it somewhere?\n\n\n-- \n------------\npasman\n",
"msg_date": "Mon, 10 Oct 2011 16:27:55 +0200",
"msg_from": "=?... |
[
{
"msg_contents": "I talked to Robert Haas and he said that index-only scans do not\noptimize COUNT(*). Is this something we can do for PG 9.2? Is anyone\nworking on this?\n\n-- \n Bruce Momjian <bruce@momjian.us> http://momjian.us\n EnterpriseDB http://enterprisedb.com\n... |
[
{
"msg_contents": "I'm looking into upgrading a fairly busy system to 9.1. They use\nserializable mode for a few certain things, and suffer through some\nserialization errors as a result. While looking over the new\nserializable/SSI documentation, one thing that stood out is:\n\nhttp://www.postgresql.org/docs/c... |
[
{
"msg_contents": "2011/10/10 Robert Haas <robertmhaas@gmail.com>:\n> On Mon, Oct 10, 2011 at 1:36 PM, Kevin Grittner\n> <Kevin.Grittner@wicourts.gov> wrote:\n>>\n>> That gives you an index-only scan; but without the WHERE clause it\n>> uses a seq scan. I think it's mainly a matter of doing enough\n>> benchmar... |
[
{
"msg_contents": "It occurs to me that we could really use two things to make it easier to\nmove copies of database stuff around:\n\npg_dump -r, which would include a CREATE ROLE for all roles needed to\nrestore the database (probably without passwords), and\n\npg_dumpall -r --no-passwords which would dump the... |
[
{
"msg_contents": "> Jeff Janes wrote:\n> Kevin Grittner wrote:\n \n>> create table t (id int not null primary key);\n>> insert into t select generate_series(1, 1000000);\n>> vacuum freeze analyze;\n>> explain analyze select count(*) from t\n>> where id between 500000 and 500010;\n>>\n>> That gives you an ind... |
[
{
"msg_contents": "Hello\n\nA current limits of dynamic work with row types in PL/pgSQL can be\ndecreased with a possible casts between rows and arrays. Now we have a\nlot of tools for arrays, and these tools should be used for rows too.\n\npostgres=# \\d mypoint\nComposite type \"public.mypoint\"\n Column │ T... |
[
{
"msg_contents": "Hi all,\n\nI wonder, could the recent work on index only scans pave the way for auto clustered tables? Consider a wide, mostly insert table with some subset of columns that I'd like to cluster on. I'm after locality of tuples that are very frequently fetched together, but not keen on the do... |
[
{
"msg_contents": "The reported git failures from the buildfarm should clear on the next run -\nthey're the result of an upgrade to the git package on the git server which\nleft the git-daemon in an non-functioning state. It's back up now.\n\nApologies for the noise.\n\n-- \nDave Page\nBlog: http://pgsnake.blog... |
[
{
"msg_contents": "I was wondering if the index-only-scan will be available in 9.2? This is not having to visit the real data to answer a query, if all the information is available in the index. I think this will be a mayor step in overtaking the big O in the g-spot.\n\nThere seams to already be some patch for ... |
[
{
"msg_contents": "Hello all,\n\nIn https://launchpad.net/bugs/835502 it was reported that the 9.1.1\ncontrib *.sql files contain the token \"MODULE_PATHNAME\", which is\nunknown:\n\n psql test < /usr/share/postgresql/9.1/extension/intarray--1.0.sql\n\nThis fails with a ton of errors about the file \"MODULE_PA... |
[
{
"msg_contents": "Hi,\n\nAn example in the PG documentation gives an error:\n\nhttp://www.postgresql.org/docs/9.1/interactive/plperl-global.html\n\nCREATE OR REPLACE FUNCTION myfuncs() RETURNS void AS $$\n $_SHARED{myquote} = sub {\n my $arg = shift;\n $arg =~ s/(['\\\\])/\\\\$1/g;\n re... |
[
{
"msg_contents": "Hi,\n\nIn 9.2dev and 9.1, when walreceiver detects an error while sending data to\nWAL stream, it always emits ERROR even if there are data available in the\nreceive buffer. This might lead to loss of transactions because such\nremaining data are not received by walreceiver :(\n\nTo prevent t... |
[
{
"msg_contents": "On Sun, Sep 11, 2011 at 10:11 AM, Josh Kupershmidt <schmiddy@gmail.com> wrote:\n> On Sat, Sep 10, 2011 at 7:47 PM, Thom Brown <thom@linux.com> wrote:\n>> Just tested this out on current master. I tried this on every object\n>> capable of having a comment, and the view reports all of them wit... |
[
{
"msg_contents": "I use rss to follow up on patches that I'm interested in and it's the \nsecond time I was wonering where my patch has gone in the commitfest app \ndue to $Topic.\n\nIs this a known limitation?\nIf yes: Is there a way to change this?\nIf yes: Can/shall I help?\nIf yes: Where should I start?\n... |
[
{
"msg_contents": "I've been thinking about the terminology used for various tools and\nextensions in PostgreSQL. The documentation bundles all them together\nin \"Additional Supplied Modules, but really some are tools, some are\nlibraries and some are extensions:\n\nTools:\noid2name\npg_archivecleanup\npgbenc... |
[
{
"msg_contents": "I noticed a problem of get_object_address() with missing_ok = true.\n\nWhen we try to solve the name of nonexistent rule/trigger/constraint on\na particular existing table, get_object_address_relobject() opens the\nrelation, but address.objectId = InvalidOid shall be set.\n\nI think it should... |
[
{
"msg_contents": "Hi,\n\nWe have several users working on a 8.4 database, using it as a\nback-end for several related apps and transfering data to and from it.\nThe database tends to get a bit messy, so i've made a little table to\nprovide an overview.\nThis table is truncated and refilled daily, it shows all ... |
[
{
"msg_contents": "Hi,\n\nI just noticed that the VACUUM process touches a lot of relations\n(affects mtime) but for one file I looked at, it didn't change. This\ndoesn't always happen, and many relations aren't touched at all.\n\nI had the following relation:\n\n-rw------- 1 thom staff 40960 13 Oct 16:... |
[
{
"msg_contents": "Hello,\n\nI am writing an extension to easily execute queries with conditions\nexpressing constraints in fuzzy logics.\n\nI wrote some C functions that get or set global variables in C.\nThe variables are MU (FLOAT : degree of a fuzzy predicate), ALPHA\n(FLOAT : threshold for filtering predic... |
[
{
"msg_contents": "Consider this example in an empty database:\n\ndb=# create table t1 (f1 int);\nCREATE TABLE\ndb=# create unique index t1f1 on t1(f1);\nCREATE INDEX\ndb=# create table t2 (f2 int references t1(f1));\nCREATE TABLE\ndb=# create table t3(f3 int primary key);\nNOTICE: CREATE TABLE / PRIMARY KEY w... |
[
{
"msg_contents": "All,\n\nI'm noticing some inconsistent and (I believe) undesirable behavior on\nRAISE INFO.\n\nIf you call a function, and it posts progress reports using RAISE INFO,\nthen you get the INFO statements plain back to the client. However, if\nthat function calls another function, then you also ... |
[
{
"msg_contents": "So I'm testing a patch to make the planner use measured all-visible-page\ncounts for index-only scans. I was expecting to possibly see some plan\nchanges in the regression tests, but this diff scared me:\n\n***************\n*** 906,921 ****\n FROM tenk1 WHERE unique1 < 10;\n sum | unique1... |
[
{
"msg_contents": "Almost immediately after we committed index-only scans, there were\ncomplaints that it didn't work with \"indexkey IN (...)\" conditions, that\nis ScalarArrayOpExpr index quals. That's because the current code only\nsupports ScalarArrayOpExpr as a bitmap indexscan qual, not a regular\nindexs... |
[
{
"msg_contents": "Hi,\n\nattached is a patch implementing the usage of SPI cursors in PL/Python.\nCurrently when trying to process a large table in PL/Python you have\nslurp it all into memory (that's what plpy.execute does).\n\nThis patch allows reading the result set in smaller chunks, using a SPI\ncursor be... |
[
{
"msg_contents": "> > Linas, could you capture the output of pg_controldata *and* increase the\n> > log level to DEBUG1 on the standby? We should then see nextXid value of\n> > the checkpoint the recovery is starting from.\n>\n> I'll try to do that whenever I'm in that territory again... Incidentally,\n> recen... |
[
{
"msg_contents": "Hi,\n\nthis idea has cropped up last PGCon - the ability to set GUC variables\nfor the duration of a single query. It would work by setting the GUCs\nfor the duration of the query and setting them back to what they were\nafter it has terminated. By \"setting them back\" I mean respecting the\... |
[
{
"msg_contents": "On master, I see a minor test error (at least on my machine) as well as\na diff. Patch attached.\n\nRegards,\n\tJeff Davis",
"msg_date": "Sun, 16 Oct 2011 10:51:09 -0700",
"msg_from": "Jeff Davis <pgsql@j-davis.com>",
"msg_from_op": true,
"msg_subject": "(patch) regression dif... |
[
{
"msg_contents": "Hi,\n\nIn 9.1, the following procedure writes a junk message into the server log;\n\n$ initdb -D data\n$ cat << EOF >> data/postgresql.conf\nwal_level = hot_standby\nmax_wal_senders = 1\nsynchronous_standby_names = '*'\nEOF\n$ pg_ctl -D data start\n$ psql -c \"create table t (i int)\" &\n$ pg... |
[
{
"msg_contents": "I just noticed that HeapTupleHeaderAdvanceLatestRemovedXid is comparing Xmax as a TransactionId without verifying whether it is a multixact or not. Since they advance separately, this could lead to bogus answers. This probably needs to be fixed. I didn't look into past releases to see if t... |
[
{
"msg_contents": "Hello there,\n two guys of our developer team ( Lorenzo and Federico )\nhave seen a strange behaviour (in 8.4 and 9.1.1 ) on update, and I think is\na bug or something really strange or I not understand correctly this\nbehavior .\n\nI explain now ( begin transaction or auto ... |
[
{
"msg_contents": "I'm not sure if these can/should be fixed or not, but here are the\ncompiler warnings I'm getting on gcc and clang on ubuntu 11.10 with -O2.\nThe gcc ones are mostly new.\n\n==================== GCC ========================\n\n$ gcc --version\ngcc (Ubuntu/Linaro 4.6.1-9ubuntu3) 4.6.1\nCopyrig... |
[
{
"msg_contents": "Hi,\n\nI noticed that if the hba_file setting in the config is uncommented\nand set to a directory instead of the full path to the file, no error\noccurs when the service starts.\n\nFor example:\n\nhba_file = '/home/thom/Development/data'\n\nThe problem with this is you cannot get into the da... |
[
{
"msg_contents": "Jeff Davis wrote:\n \n> I'm not sure if these can/should be fixed or not, but here are the\n> compiler warnings I'm getting on gcc and clang on ubuntu 11.10 with\n> -O2.\n \n> elog.c: In function ‘write_pipe_chunks’:\n> elog.c:2479:8: warning: ignoring return value of ‘write’, declared\n> wi... |
[
{
"msg_contents": "Hi list,\n\nPostgreSQL 9.0 introduced the \"join removal\" feature for cases where\nan left join can't return more than one output row. This feature\nrelies on checking whether a UNIQUE constraint exists on the joined\ncolumn in the referenced table.\n\nHowever, there was another new feature ... |
[
{
"msg_contents": "Adding CORRESPONDING to Set Operations\nInitial patch, filename: corresponding_clause_v2.patch\n\nThis patch adds CORRESPONDING clause to set operations according to\nSQL20nn standard draft as Feature F301, \"CORRESPONDING in query\nexpressions\"\n\nCorresponding clause either contains a BY(.... |
[
{
"msg_contents": "Is it really necessary for GetSerializableTransactionSnapshotInt to\nacquire an empty SERIALIZABLEXACT before it acquires a snapshot?\nIf so, why? The proposed synchronized-snapshots feature will mean\nthat the allegedly-new snapshot actually was taken some time before,\nso it seems to me th... |
[
{
"msg_contents": "Hackers,\n\nWe've just found an issue with pg_dumpall in 9.1.1 where a dump starts with lines like these:\n\n ALTER ROLE dude WITH NOSUPERUSER INHERIT NOCREATEROLE NOCREATEDB LOGIN PASSWORD 'md5bdd7f8e73a214981b1519212b02a5530' VALID UNTIL 'infinity';\n ALTER ROLE dude SET default_table... |
[
{
"msg_contents": "ProcessStandbyHSFeedbackMessage has a race condition: it thinks it can\ncall GetOldestXmin and then the result will politely hold still while\nit considers what to do next. But in fact, whoever has the oldest xmin\ncould exit their transaction, allowing the global minimum to advance.\nIf a V... |
[
{
"msg_contents": "I've recently gotten annoyed again by the sorry state of docbook\nSGML support in macports, and finally decided to do something\nabout it (OK, the fact that I recently deleted my Ubuntu VM,\nforgetting that one of the reasons I had it was to be able to\nbuild our docs has something to do with... |
[
{
"msg_contents": "Dan Ports wrote:\n \n> I think it would be fairly sensible to push some of this into\n> ProcArray, actually. The commit sequence numbers just involve\n> assigning/incrementing a global counter when taking a snapshot and\n> finishing a transaction -- that's not too much work, doesn't\n> requi... |
[
{
"msg_contents": "\nI just noticed that DropTrigger uses ShareRowExclusiveLock to lock the\nrelation it's going to drop the trigger on. The comment right above it\nsays that this should match RemoveTriggerById, but that one uses\nAccessExclusiveLock -- so one of them (or the comment) is not right.\n\nBesides,... |
[
{
"msg_contents": "It looks like \\copy is just passing the text of the query unadjusted to \n\"COPY\". I get a syntax error on \":x\" with the \\copy below on both 9.0 \nand 9.1\n\n=== test script ===\n\\set x '''HELLO'''\n-- Works\n\\echo :x\n-- Works\n\\o '/tmp/test1.txt'\nCOPY (SELECT :x) TO STDOUT;\n-- Doe... |
[
{
"msg_contents": "I've thought of another nasty problem for the sync-snapshots patch.\nConsider the following sequence of events:\n\n1. Transaction A, which is about to export a snapshot, is running in\n database X.\n2. Transaction B is making some changes in database Y.\n3. A takes and exports a snapshot sh... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.