threads listlengths 1 2.99k |
|---|
[
{
"msg_contents": "Hi hackers,\nPlease notice that it is not a first of April joke;)\n\nSeveral times we and our customers are suffered from the problem that \nPostgres got stuck in deadlock detection.\nOne scenario is YCSB workload with Zipf's distribution when many clients \nare trying to update the same reco... |
[
{
"msg_contents": "Hello, hackers!\n\nWhen debugging is enabled for server logging, isolation tests fail \nbecause there're no corresponding output functions for InsertStmt / \nDeleteStmt / UpdateStmt that are used in the output of the MergeAction \nnodes (see the attached regressions diffs and output). I also ... |
[
{
"msg_contents": "Foreign keys on partitioned tables\n\nAuthor: Álvaro Herrera\nDiscussion: https://postgr.es/m/20171231194359.cvojcour423ulha4@alvherre.pgsql\nReviewed-by: Peter Eisentraut\n\nBranch\n------\nmaster\n\nDetails\n-------\nhttps://git.postgresql.org/pg/commitdiff/3de241dba86f3dd000434f70aebba725f... |
[
{
"msg_contents": "Hi,\n\nWhile looking at resolving [1] I re-read heap_lock_tuple() and\nsubsidiary routines and got thoroughly confused for a while.\n\nOne reason was that function names and comments talk about updated, when\nthey also actually deal with deletes. heap_lock_updated_tuple()\nspecifically is ca... |
[
{
"msg_contents": "Hi,\n\nhttps://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=spurfowl&dt=2018-04-05%2003%3A22%3A05\n\n*** /home/pgbuild/buildfarm/HEAD/pgsql.build/src/test/regress/expected/select_parallel.out\n2018-04-04 23:22:07.297108347 -0400\n--- /home/pgbuild/buildfarm/HEAD/pgsql.build/src/test/regres... |
[
{
"msg_contents": "Hello, for the support of multiple versions of PostgreSQL RPM packages on\none system, we are thinking about having only one libpq.so.5\n(libecpg.so.6, libpgtype.so.3 respectively) supported and about building\n(linking) all the PostgreSQL package versions against that. The pattern\nwould me... |
[
{
"msg_contents": "I started looking at the \"Improve compactify_tuples and \nPageRepairFragmentation\" patch, and set up a little performance test of \nWAL replay. I ran pgbench, scale 5, to generate about 1 GB of WAL, and \ntimed how long it takes to replay that WAL. To focus purely on CPU \noverhead, I kept ... |
[
{
"msg_contents": "Hello,\n\nI want to get the target Relation name for a UPDATE / INSERT / DELETE in a planner_hook. Do I understand struct Query correctly that:\n\nQuery->resultRelation will be the index into Query->rtable to give me the target Relation?\n\nAnd if yes, what would rtable give me as list entry?... |
[
{
"msg_contents": "Hi,\n\nDoes anyone know, why `typcategory` value for tsvector `regconfig` is\n`TYPCATEGORY_NUMERIC`, but in all the tests it's being used in string format?\nIt's probably not a big deal, but in this thread [1] it prevents me from\nadopting the nice solution with a boolean flag for `to_tsvecto... |
[
{
"msg_contents": "Hi,\n\nI wanted to share the project I've been working on which dynamically rewrites queries to target materialized views when views are available that can satisfy part of a query with lower cost plans.\n\nI embarked upon as an interesting side project. It took me a bit more time than I antic... |
[
{
"msg_contents": "This is only a preliminary report, I'm still trying to analyze what's\ngoing on, but:\n\nIn doing testing on FreeBSD with a filesystem set up to induce errors\ncontrollably (using gconcat+gnop), I can get this to happen (on 11devel):\n\n(note that \"mytable\" is on a tablespace on the errorin... |
[
{
"msg_contents": "Hi,\n\nI have a tif image of 483 mb, i am uploading the raster into postgis using\n\n\"*raster2pgsql -I -C\n /home/XXXXXX/Downloads/20161207T055222_20161207T055323_T42QYL/20161207T055222_20161207T055323_T42QYL.ndvi.tif\npostgis.gistesttable6 |psql -h localhost -U XXXX -d gisdb -p 5432*\"\n\ni... |
[
{
"msg_contents": "Hello David,\r\n\r\nactually Query->resultRelation is one based and Query->rtable is zero based.\r\n\r\nSo the correct call to list_nth is: list_nth(Query->rtable, Query->resultRelation - 1)\r\n\r\nThe comment \"rtable index of target relation for INSERT/UPDATE/DELETE/MERGE; 0 for SELECT\" on... |
[
{
"msg_contents": "On Thu, 2018-03-22 at 11:16 -0700, legrand legrand wrote:\n> Reading other pg_stat_statements threads on this forum, there are\n> also activ\n> developments to add:\n> - planing duration,\n> - first date,\n> - last_update date,\n\nAs I see it, planning duration, first date, and last update da... |
[
{
"msg_contents": "Those of you who've been paying attention to the bootstrap data conversion\nthread will know that one of the key ideas is to put everything in the\ncatalog headers that's of direct use to client-side code into separate\n\"pg_foo_d.h\" headers. This allows clients to include pg_foo_d.h to get... |
[
{
"msg_contents": "My patch is hidden in a very long thread, so I thought letting more\nwidely known that I'm proposing to add a new subdir \"partitioning\" under\nsrc/backend and src/include.\nTBH I think the new idea is better than our current arrangement where we\nhave catalog/partition.c which is a dumping ... |
[
{
"msg_contents": "I felt it was worth spending some extra effort on documentation for\nthis change, since it's going to impact a lot of future patches.\nAccordingly, I've taken John's proposed README text and moved it\ninto the SGML format, and done a fair amount of editing to extend\nthe text and bring it all... |
[
{
"msg_contents": "It sure looks like there's been a frantic push to commit stuff that\nmaybe wasn't quite fully baked. I'm not terribly on board with that,\nbecause it's likely to be hard to disentangle who broke what.\nBut in particular, it's clear that partition_prune and\nisolation/checksum_cancel are show... |
[
{
"msg_contents": "Allow insert and update tuple routing and COPY for foreign tables.\n\nAlso enable this for postgres_fdw.\n\nEtsuro Fujita, based on an earlier patch by Amit Langote. The larger\npatch series of which this is a part has been reviewed by Amit\nLangote, David Fetter, Maksim Milyutin, Álvaro Herr... |
[
{
"msg_contents": "Hi,\n\nAs Daniel pointed out in:\nhttps://postgr.es/m/FB948276-7B32-4B77-83E6-D00167F8EEB4@yesql.se the\npg_atomic_flag fallback implementation is broken. That has gone\nunnoticed because the fallback implementation wasn't testable until now:\n- /* ---\n- * Can't run the test under the ... |
[
{
"msg_contents": "Hasn't been updated as built-in support grew for temp files, DSM segments,\nand JIT contexts.\n\nWith enough luck, a README update won't break anything.\n\n-Chap",
"msg_date": "Fri, 6 Apr 2018 19:46:25 -0400",
"msg_from": "Chapman Flack <chap@anastigmatix.net>",
"msg_from_op": tru... |
[
{
"msg_contents": "Hi,\n\nsqlsmith found a query that triggers the following assertion in master\nas of 039eb6e92f:\n\nTRAP: FailedAssertion(\"!(subpath->parallel_safe)\", File: \"pathnode.c\", Line: 1813)\n\nBacktrace and recipe against the regression database below.\n\nregards,\nAndreas\n\n#0 __GI_raise (sig... |
[
{
"msg_contents": "Hi,\n\ntesting with master at 039eb6e92f yielded another query triggering an\nassertion. Backtrace and query against the regression database below.\n\nregards,\nAndreas\n\n#0 __GI_raise (sig=sig@entry=6) at ../sysdeps/unix/sysv/linux/raise.c:51\n#1 0x00007f25474cf42a in __GI_abort () at ab... |
[
{
"msg_contents": "Hi,\n\nthe following query triggers a segfault for me when run against the\nregression database. Testing was done with master at 039eb6e92f.\nBacktrace below.\n\nregards,\nAndreas\n\nselect\n case when pg_catalog.lastval() < pg_catalog.pg_stat_get_bgwriter_maxwritten_clean() then case when ... |
[
{
"msg_contents": "Support partition pruning at execution time\n\nExisting partition pruning is only able to work at plan time, for query\nquals that appear in the parsed query. This is good but limiting, as\nthere can be parameters that appear later that can be usefully used to\nfurther prune partitions.\n\nT... |
[
{
"msg_contents": "Hi,\n\nIt's common that half the buildfarm has reported back before a single\nwindows buildfarm animal reports. And if they report a failure one often\nhas to wait for hours for the next run.\n\nIt'd be awesome if somebody could set up a windows animal that runs\nfrequently (i.e. checks for b... |
[
{
"msg_contents": "Hello,\n\nI've discovered a couple of bugs in logical decoding code, both leading\nto incorrect decoding results in somewhat rare cases. First, xmin of\nslots is advanced too early. This affects the results only when\ninterlocking allows to perform DDL concurrently with looking at the\nschema... |
[
{
"msg_contents": "(This was discussed in the \"way forward on bootstrap data\" thread,\nbut we didn't do anything about it.)\n\nTraditionally genbki.pl has printed \"Writing foo\" for every file\nit writes out. As of 372728b0d that is a heck of a lot more files\nthan it once was, so you get this:\n\n$ make -j... |
[
{
"msg_contents": "Merge catalog/pg_foo_fn.h headers back into pg_foo.h headers.\n\nTraditionally, include/catalog/pg_foo.h contains extern declarations\nfor functions in backend/catalog/pg_foo.c, in addition to its function\nas the authoritative definition of the pg_foo catalog's rowtype.\nIn some cases, we'd ... |
[
{
"msg_contents": "Hi all,\n\nWhile looking at the output of the TAP tests, I have seen warnings like\nthe following:\nUse of uninitialized value $target_lsn in concatenation (.) or string at\n/home/foo/git/postgres/src/bin/pg_rewind/../../../src/test/perl/PostgresNode.pm\nline 1540.\n[...]\n./src/bin/pg_baseba... |
[
{
"msg_contents": "Hi all,\n\nI was just going through pg_rewind's code, and noticed the following\npearl:\n /*\n * Don't allow pg_rewind to be run as root, to avoid overwriting the\n * ownership of files in the data directory. We need only check for root\n * -- any other user won't have sufficie... |
[
{
"msg_contents": "Hi.\n\nI noticed that the newly added pruning does not work if the partition key\nis of one of the types that have a corresponding pseudo-type.\n\n-- array type list partition key\ncreate table arrpart (a int[]) partition by list (a);\ncreate table arrpart1 partition of arrpart for values in ... |
[
{
"msg_contents": "While I was looking on some patch, I happened to notice that\nthere's a trivial typo in rewriteManip.c\n\n> a ConvertRowTypeExpr to map back to the rowtype expected by the expression.\n\nThe correct name for \"ConvertRowTypeExpr\" is \"ConvertRowtypeExpr\"\n\nregards,\n\n-- \nKyotaro Horiguch... |
[
{
"msg_contents": "Hi hackers,\n\nPostgres optimizer is not able to build efficient execution plan for the \nfollowing query:\n\nexplain select * from people_raw where not (\"ID\"<2068113880 AND \"INN\" \nis not null) and \"ID\"<=2068629726 AND \"INN\" is not null;\n QU... |
[
{
"msg_contents": "Revert \"Allow on-line enabling and disabling of data checksums\"\n\nThis reverts the backend sides of commit 1fde38beaa0c3e66c340efc7cc0dc272d6254bb0.\nI have, at least for now, left the pg_verify_checksums tool in place, as\nthis tool can be very valuable without the rest of the patch as we... |
[
{
"msg_contents": "Hi all,\n\nI have not been giving much attention to the thread about enabling\nchecksums online, which has resulted in the revert of the feature, but\nthere is still pg_verify_checksums around. So I looked at it a bit.\n\nI have a couple of questions/gotchas about it:\n1) The documentation s... |
[
{
"msg_contents": "Hi,\n\nTrying covering indexes on partitioned tables i get this error\n\"\"\"\npostgres=# create index on t1_part (i) include (t);\nERROR: cache lookup failed for opclass 0\n\"\"\"\n\nTo reproduce:\n\ncreate table t1_part (i int, t text) partition by hash (i);\ncreate table t1_part_0 partiti... |
[
{
"msg_contents": "Hi.\n\nI noticed that RelationBuildPartitionKey() is not doing the right thing\nwith respect to which context it passes to fmgr.c to set a the (cached)\npartition support function's FmgrInfo's fn_mcxt.\n\nI noticed a crash while trying to change partition pruning tests to use\nmanually create... |
[
{
"msg_contents": "Hi,\n\nThere are some cases that power() function does not work \ncorrectly with 'NaN' arguments in Windows environment.\nSomething like,\n\npostgres=# select power('NaN',11);\nERROR: value out of range: underflow\npostgres=# select power('NaN','NaN');\nERROR: value out of range: underflow\... |
[
{
"msg_contents": "Hi,\n\nI used to run\n\n\"\ncd src/backend\nmake submake-errcodes\n\"\n\nin the RPM spec file, but looks like it was removed recently. Is that replaced\nwith something else, or removed completely?\n\nRegards,\n-- \nDevrim Gündüz\nEnterpriseDB: https://www.enterprisedb.com\nPostgreSQL Consulta... |
[
{
"msg_contents": "Hello,\n\nOne of our 2ndQuadrant support customers recently reported a sudden rush of\nTOAST errors post a crash recovery, nearly causing an outage. Most errors\nread like this:\n\nERROR: unexpected chunk number 0 (expected 1) for toast value nnnn\n\nWhile we could bring back the cluster to n... |
[
{
"msg_contents": "Hackers,\n\nI have gone through all the remaining non-bug entries in CF 2018-03 and\npushed them or closed them as appropriate. The majority of the patches\nwere in Needs Review or Ready for Committer status and I did a brief\nreview of each to be sure the state was reasonable.\n\nI moved a ... |
[
{
"msg_contents": "In writing extension update scripts, I find it to be really difficult to\ngrok diffs for example in changed view or function definitions when a new\nextension script has to include the whole definition in a new file. I want\nto rather use separate files for these objects, then use something ... |
[
{
"msg_contents": "A report from a pgbouncer user revealed that running pg_dump -C/--create\ndoes not work through a connection proxy if the virtual database name on\nthe proxy does not match the real database name on the database server.\nThat's because pg_dump looks up the database to be dumped using the\ninf... |
[
{
"msg_contents": "Hi,\n\nToday looking for information on hard limits for\nautovacuum_vacuum_cost_limit I found myself with a very short\ndescription in the docs.\n\nThis is a patch to add some further description, plus the upper and\nlower limits it has.\n\n-- \nMartín Marqués\nselect 'martin.marques' || '@' ... |
[
{
"msg_contents": "Hi all,\n\nI am catching up with new features so I have begun going through\nCovering indexes. While reading the code, I have noticed a couple of\nthings:\n1) Some typos.\n2) An inconsistent variable name in pg_dump.\n\nThe thing and structure of the code is pretty interesting by the way.\nT... |
[
{
"msg_contents": "Hi all,\n\nThe documentation of covering indexes is incorrect for CREATE and ALTER\nTABLE:\n- ALTER TABLE's page is missing the call.\n- Exclusion constraints can use INCLUDE clauses.\n\nIn order to simplify the documentation, please let me suggest the\nattached which moves the definition of ... |
[
{
"msg_contents": "I have discovered that:\n\n\tmake clean; make check\n\nfails with:\n\n\t/bin/mkdir -p '/pgtop'/tmp_install/log\n\tmake -C '.' DESTDIR='/pgtop'/tmp_install install >'/pgtop'/tmp_install/log/install.log 2>&1\n\tsrc/Makefile.global:388: recipe for target 'temp-install' failed\n\tmake: *** [temp-... |
[
{
"msg_contents": "Just had someone report that pg_partman wasn't handling tablespaces for\nnative partitioning.\n\nhttps://github.com/keithf4/pg_partman/issues/212\n\nI'd assumed that that was a property that was being inherited from the\nparent table, but apparently the TABLESPACE flag to CREATE TABLE is\ncom... |
[
{
"msg_contents": "Hi,\n\npsql (PostgreSQL) 10.3\n\nHere are the steps to reproduce a leak:\n\n1. connect to 10.3 server, perform the query similar to:\n\nselect 'message' || generate_series(1,1000000000);\n\n2. monitoring psql memory usage in htop or similar tool, press ctrl+c at\nsome point where you can clea... |
[
{
"msg_contents": "Hi, hackers!\n\nLooks like we finally have covering indexes! And that's cool!\n\nSo I decided to create a thread to discuss covering GiST indexes.\nHere's a prototype patch implementing this functionality.\nIt is quite small (+80 -30) and lacks tests and docs. But it creates a context.\n\nI h... |
[
{
"msg_contents": "We've been seeing $subject since commit 1bc0100d2, with little clue\nas to the cause. Previous attempts to fix it by preventing autovacuum\nfrom running on the relevant tables didn't seem to help.\n\nI have now managed to reproduce the issue reliably enough to study it.\n(It turns out that o... |
[
{
"msg_contents": "Running \"make installcheck\" with wal_consistency_checking='all' on the\nmaster branch shows the follow failure on a streaming replica:\n\n19696/2018-04-12 11:35:29 PDT FATAL: inconsistent page found, rel\n1663/50192/66636, forknum 0, blkno 0\n19696/2018-04-12 11:35:29 PDT CONTEXT: WAL red... |
[
{
"msg_contents": "I recently ran into an issue where (after implementing fulltext search on\nmy site) a user searching real estate listings for \"no pets\" also got\nresults for \"pets OK\"! This was obviously a problem. After investigating,\nit seems the word \"no\" is considered a stopword by default (it's i... |
[
{
"msg_contents": "Hi,\n\nApologies if this was already discussed, I didn't see it. One of my\nanimals elver had a one-off failure this morning:\n\nhttps://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=elver&dt=2018-04-12%2018%3A18%3A05\n\n partition_prune ... FAILED\n\n ... |
[
{
"msg_contents": "itup.h says of MinIndexTupleSize/MaxIndexTuplesPerPage:\n\n/*\n * MaxIndexTuplesPerPage is an upper bound on the number of tuples that can\n * fit on one index page. An index tuple must have either data or a null\n * bitmap, so we can safely assume it's at least 1 byte bigger than a bare\n *... |
[
{
"msg_contents": "Hi all,\n\nI was just playing with the WAL consistency issue with rows moved across\npartitions when I noticed that heapdesc.c is not really talkative about\nthe different flag records set.\n\nWhat about something like the patch attached? I found that useful for\ndebugging.\n\n(One comment o... |
[
{
"msg_contents": "Hello hackers,\n\nCan you please explain, is this a bug or intended behaviour?\n\nRunning as non-privileged user:\n\npostgres=> SELECT datid, datname FROM pg_stat_database FOR UPDATE;\nERROR: permission denied for view pg_stat_database (SQLState: 42501)\n\nBut:\n\npostgres=> CREATE VIEW pgsd ... |
[
{
"msg_contents": "Andres Freund wrote:\n\n> It'd be good to adjust the thread topic - this surely isn't about the\n> crash anymore. And it's good, especially given we're past the feature\n> freeze etc, if the subject conveyed what's happening?\n\nSure thing. This thread comes from:\nhttps://postgr.es/m/20180... |
[
{
"msg_contents": "*Hello,*\n\nI'm new here. I'm David and would describe myself as an ambitious newbie,\nso please take my suggestion with a grain of salt.\n\n*Use case:*\nI find it difficult to properly parse postgres logs into some kind of log\naggregator (I use fluent bit). My two standard option are standa... |
[
{
"msg_contents": "Currently, we don't set rpath for llvmjit.so to include `llvm-config --\nlibdir`. That means you need to keep LD_LIBRARY_PATH set properly to\nuse any non-system installation of LLVM.\n\nThis isn't unique to LLVM (we only set rpath to the postgres $libdir),\nso I thought it might be project p... |
[
{
"msg_contents": "Andreas Seltenreich's sqlsmith tool has found an impressive number of\nbugs. In light of that, it seems to me that it would be reasonable for\na contributor to write a regression test that avoids dropping database\nobjects specifically so that sqlsmith had some chance of finding bugs\nafter t... |
[
{
"msg_contents": "Hi\n\nI am testing PostgreSQL in customer's environment and I found some\nunexpected behave.\n\nWhen auto_explain is used, and there is slower parallel query, then\nauto_explain raises log for any slow process.\n\n arallel worker for PID 20089 : auto_explain database=NULL client=local\nappn... |
[
{
"msg_contents": "Hi,\n\nmy customer does performance checks of PostgreSQL 9.5 and 10. Almost all\nqueries on 10 are faster, but there are few queries (40 from 1000) where\nPostgreSQL 9.5 is significantly faster than. Unfortunately - pretty fast\nqueries (about 20ms) are too slow now (5 sec).\n\nattached execu... |
[
{
"msg_contents": "Hi folks,\n\nI got some complex query which works on PostgreSQL 9.6 , but fails on\nPostgreSQL 10.\n\nVersion of PostgreSQL:\nPostgreSQL 10.3 on x86_64-apple-darwin14.5.0, compiled by Apple LLVM\nversion 7.0.0 (clang-700.1.76), 64-bit\n\nSimplified core of the problematic query looks like thi... |
[
{
"msg_contents": "Hi,\n\nWhile updating the fix-postgres_fdw-WCO-handling patch, I noticed that\nthe patch for tuple routing for foreign partitions doesn't work well\nwith remote triggers. Here is an example:\n\npostgres=# create table loct1 (a int check (a in (1)), b text);\npostgres=# create function br_ins... |
[
{
"msg_contents": "Hi,\nConsider this scenario\n\npostgres=# CREATE TABLE plt (a int, b int) PARTITION BY LIST(a);\npostgres=# CREATE TABLE plt_p1 PARTITION OF plt FOR VALUES IN (1);\npostgres=# CREATE TABLE plt_p2 PARTITION OF plt FOR VALUES IN (2);\npostgres=# INSERT INTO plt VALUES (1, 1), (2, 2);\npostgres=... |
[
{
"msg_contents": "As reported in [1], pg_attribute.attndims turns to 0 when 'create table \nlike/as'.\nThe patch attached is to fix it.\n\nBest Regards,\n Alexey\n\n[1] \nhttps://www.postgresql.org/message-id/20150707072942.1186.98151%40wrigleys.postgresql.org",
"msg_date": "Mon, 16 Apr 2018 15:01:50 +01... |
[
{
"msg_contents": "Hi,\n\nAn issue [1] reported that pg_recvlogical emitted an error in 9.6.8. I\nconfirmed that it was broken in recent minor versions (9.6.8, 9.5.12,\n9.4.17 -- using same server/client version). It was broken by commit\n582edc369cdbd348d68441fc50fa26a84afd0c1a and its siblings.\n\n$ postgres ... |
[
{
"msg_contents": "The following bug has been logged on the website:\n\nBug reference: 15160\nLogged by: Alexey Ermakov\nEmail address: alexey.ermakov@dataegret.com\nPostgreSQL version: 10.3\nOperating system: Linux\nDescription: \n\nHello,\r\n\r\n\r\nI'm wondering how planner estima... |
[
{
"msg_contents": "Alvaro pinged me off list to mention the coverage tool indicates that\nthere's no test coverage for mark_invalid_subplans_as_finished() [1].\n\nThe attached patch aims to put that right. This function would only\never be called during a Parallel Append in a query which has some exec\nParams b... |
[
{
"msg_contents": "For the bootstrap data conversion, it was desirable for postgres.bki\nto remain unchanged, so some ugly quoting hacks were added to\ngenbki.pl to match the quoting conventions in the DATA() lines. At\nthis point, it's possible (and worthwhile I think) to remove those,\nand along the way simpl... |
[
{
"msg_contents": "Hello,\n\nPG-Strom v2.0 is released. It is an extension module for PostgreSQL\nto accelerate reporting and analytics workloads towards large scale\ndata set using GPU.\n\nMajor enhancement in PG-Strom v2.0 includes:\n\n- Overall redesign of the internal infrastructure to manage GPU and\nstabi... |
[
{
"msg_contents": "Now that feature freeze is past, I wonder if it's time to run pgindent.\n\nLast year we did a run immediately after beta1, plus one just before\nbranching off REL_10_STABLE. The value of an early run, IMO, is to\nget most of the changes in place so that people have a stable base\nto work fro... |
[
{
"msg_contents": "Hi,\n\nI'd like to propose to add $SUBJECT for performance improvement.\n\nWhen VACUUM tries to truncate the trailing empty pages, it scans shared_buffers\nto invalidate the pages-to-truncate during holding an AccessExclusive lock on\nthe relation. So if shared_buffers is huge, other transact... |
[
{
"msg_contents": "David Rowley wrote in https://postgr.es/m/CAKJS1f8o2Yd=rOP=Et3A0FWgF+gSAOkFSU6eNhnGzTPV7nN8sQ@mail.gmail.com\n\n> I've made another pass over the nodeAppend.c code and I'm unable to\n> see what might cause this, although I did discover a bug where\n> first_partial_plan is not set taking into ... |
[
{
"msg_contents": "woodlouse just showed a failure that looked a bit familiar:\n\nhttps://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=woodlouse&dt=2018-04-17%2016%3A42%3A43\n\nI trawled the buildfarm logs for similar failures, and there are several:\n\n sysname | snapshot | stage | ... |
[
{
"msg_contents": "In PG10 the planner's partition pruning could be disabled by changing\nthe constraint_exclusion GUC to off. This is still the case for PG11,\nbut only for UPDATE and DELETE queries. There is currently no way to\ndisable partition pruning for SELECT.\n\nShould we allow this?\n\nTo make this a... |
[
{
"msg_contents": "pg_check_frozen() reports corrupted VM freeze state.\r\n\r\nFound with one of my stress tests. Simplified to the repro below.\r\n\r\nThe reason for the 33 rows/pages is that I wanted to test if a 2nd vacuum freeze repaired the situation. I was confounded till I discovered SKIP_PAGES_THRESHO... |
[
{
"msg_contents": "Hi all,\n\nBack in 2012, Dan Scales, who was working on VMware Postgres, has posted\na patch aimed at removing the need of full-page writes by introducing\nthe concept of double writes using a double-write buffer approach in\norder to fix torn page problems:\nhttps://www.postgresql.org/messag... |
[
{
"msg_contents": "Hi,\nWhile working on a fix related to non-direct DML [1], I noticed that\npostgresExecForeignInsert(), postgresExecForeignUpdate() and\npostgresExecForeignDelete() functions are almost identical except that\npostgresExecForeignInsert() doesn't require ctid. The fix that I was\nworking is app... |
[
{
"msg_contents": "Hello hackers,\n\nCouple of years ago Victor Wagner presented a patch [1] that introduced \nmultiple hosts capability and also hostorder and failover_timeout \nparameters for libpq. Subsequently multi-host feature was reimplemented \nby Robert Haas and committed. Later target_session_attrs pa... |
[
{
"msg_contents": "Hi all.\n \nI don't know whether this is expected or not but I'm experiencing over 2x \nslowdown on a large query in PG-11 with JIT=on.\n \n(query is a prepared statement executed with \"explain analyze execute \nmyprepared(arg1, arg2, ..., argn)\")\n \nAfter 10 executions these are the resul... |
[
{
"msg_contents": "Hello,\n\nIt seems to me that our customer might have hit an unresolved data\ncorruption issue which is already known in this ML, but I can't figure\nout why this happened. I'd appreciate if you could give me your\nthoughts. Depending on the severity of this issue and the customer's\nreques... |
[
{
"msg_contents": "During recent review of the INCLUDE covering index patch, I pushed to\nformalize the slightly delicate assumptions that we make around how\nindex_truncate_tuple() is called. It's natural to call\nindex_truncate_tuple() during a page split, when a buffer lock is\nheld. This is what we actually... |
[
{
"msg_contents": "in commit 8561e48, _SPI_stack alloc from TopMemoryContext. But AtEOXact_SPI just set _SPI_stack = NULL. Is this a memory leak?\r\n\r\n\r\n\r\n\r\n\n\nin commit 8561e48, _SPI_stack alloc from TopMemoryContext. But AtEOXact_SPI just set _SPI_stack = NULL. Is this a memory leak?",
"msg_date"... |
[
{
"msg_contents": "Hi all,\n\nI was just testing the VACUUM truncation logic, and bumped into what\nlooks like a corrupted btree index. Here is a reproducer:\ncreate table aa (a int primary key, b bool);\ninsert into aa values (generate_series(1,1000000), false);\ncheckpoint;\nupdate aa set b = false where a >... |
[
{
"msg_contents": "There have been several discussions of replacing PG's autoconf +\nsrc/tools/msvc system. The last example is happening now at the bottom of\nthe Setting rpath on llvmjit.so thread.\n\nI see potentially big advantages to moving but also to PG's conservative\napproach that keeps it running on e... |
[
{
"msg_contents": "On Thu, Apr 19, 2018 at 11:38:09AM +0800, jian.long@i-soft.com.cn wrote:\n> in commit 8561e48, _SPI_stack alloc from TopMemoryContext. But\n> AtEOXact_SPI just set _SPI_stack = NULL. Is this a memory leak? \n\nYou are right. I can easily see the leak if I use for example a\nbackground worker... |
[
{
"msg_contents": "Hi,\n\nEncountered the following behavior with initdb on one of our test\nbuilds while using latest master head:\n\ninitdb -D $DATADIR\n\nThe program \"postgres\" was found by \"/Users/nikhils/install/bin/initdb\"\nbut was not the same version as initdb.\nCheck your installation.\n\nIntrigued... |
[
{
"msg_contents": "Various comments in GetOldestXmin mention the possibility of the oldest\nxmin going backward, and assert that this is actually safe. It's not.\n\nConsider:\n\nA table has a toastable column. A row is updated in a way that changes\nthe toasted value. There are now two row versions pointing to ... |
[
{
"msg_contents": "Hi\r\n\r\nI searching a reason why result of plpgsql_check's regress tests are\r\ndifferent when its executed from regress tests or interactive.\r\n\r\nThis is simple test\r\n\r\ndrop table testtable;\r\n\r\ncreate table testtable(a int, b int);\r\n\r\ncreate or replace function test()\r\nret... |
[
{
"msg_contents": "Hi,\n\nin MS SQL Server you are able to make an entry of SESSION_CONTEXT read-only\nby passing the @read_only param to the sp_set_session_context function:\n\"[ @read_only= ] { 0 | 1 } A flag of type bit. If 1, then the value for the\nspecified key cannot be changed again on this logical conn... |
[
{
"msg_contents": "Hi\n\nI am plpgsql_check author and now, I am solving bug related to cached\nplpgsql function. I am not able fore plpgsql compilation - and sometimes\nthe compiled version can be obsolete. As protection to false alarms, I need\nfresh compilation immediately before plpgsql_check call.\n\nI nee... |
[
{
"msg_contents": "Hi all,\n\nI have modified the PostgreSQL source code for my project implementation.\nIn the modified code, I am calling query_planner function several times\nwhile I am inside the grouping planner function.\n\nAs the number of executions increase I have noticed memory leaks in the\ncode. I h... |
[
{
"msg_contents": "bms_prev_member mistakenly has a hardcoded 24 in the function. This\nshould really be BITS_PER_BITMAPWORD - 8 so that it is properly set to\n56 if someone compiles with 64-bit bitmapwords.\n\nThe attached fixes this and also adds a test to exercise the function\na bit. [1] indicates there's c... |
[
{
"msg_contents": "Hi all,\n\nBGWORKER_BYPASS_ALLOWCONN has been added by commit eed1ce7, which is an\ninfrastructure piece to be able to enable and disable dynamically\nchecksums on a cluster. The main idea is to be able to bypass\ndatallowconn which allows a background worker to connect to a database\neven i... |
[
{
"msg_contents": "I would like to donate some code to the project, formatting numbers as any\nbase from 2 to 64. The FAQ describes contributions to the core code, but\nit's possible contrib is a better target. This is all of course contingent\non how well received this extension code is of course. Code availab... |
[
{
"msg_contents": "(patch and discussion for PG 12)\n\nOn 2018/04/22 1:28, Tom Lane wrote:\n> Amit Langote <Langote_Amit_f8@lab.ntt.co.jp> writes:\n>> [ v8-0001-Allow-generalized-expression-syntax-for-partition.patch ]\n>\n> I find what you did to a_expr here to be pretty horrid.\n\nThanks for the review.\n\n> ... |
[
{
"msg_contents": "Looks like the JIT flags definitions are not properly parenthesized.\nIt might not matter too much for their current usage, but it might\nsave a bit of confusion later.\n\ngit grep \"^#define .*<<.*[^)]$\" indicates these are the only offenders.\n\nPatch attached.\n\n-- \n David Rowley ... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.