threads listlengths 1 2.99k |
|---|
[
{
"msg_contents": "Hi,\n\nSimilar to [1] (Expression based aggregate transition / combine function\ninvocation), this patch provides small-medium performance benefits in\norder to later enable larger performance benefits with JIT compilation.\n\nWhat this basically does is to move execGrouping.c functions that\... |
[
{
"msg_contents": "Hi hackers,\n\nWhile reviewing commit c6755e23 I realised that es_query_dsa is\nbroken. It might have made some kind of sense as a name and a concept\nin an earlier version of the proposal to add a DSA area for parallel\nquery's use, when the DSA area was going to be independent of parallel\... |
[
{
"msg_contents": "The following test\n\n-- Input relation to aggregate push down hook is not safe to pushdown and thus\n-- the aggregate cannot be pushed down to foreign server.\nexplain (verbose, costs off)\nselect count(t1.c3) from ft1 t1, ft1 t2 where t1.c1 = postgres_fdw_abs(t1.c2);\n\nproduces the followi... |
[
{
"msg_contents": "I'm thinking of writing a background worker that will enumerate\nthe databases present, and spin off, for each one, another BGW\nthat will establish a connection and do stuff.\n\nFor the \"master\" one, what capabilities will it need to simply\nenumerate the current names of known databases? ... |
[
{
"msg_contents": "Hi all,\n\nSince commit 4e5fe9ad (committer Robert Haas and author Amit Langote),\ncoverity has been complaining that the new code of ExecFindPartition()\nmay use a set of values and isnull values which never get initialized.\nThis is a state which can be easily reached with the following SQL... |
[
{
"msg_contents": "Hello, hackers!\n\nI found in src/backend/utils/mb/wchar.c: pg_verify_mbstr_len() that it\nreports ASCII Null character (\\000) as invalid. As for me, it should\npass validation. However, ASCII Null character breaks a line and the\nend of the line is missed, try:\n\nINSERT INTO mytable VALUES... |
[
{
"msg_contents": "Hi hackers,\n\nThe manual implies that only Linux can use huge pages. That is not\ntrue: FreeBSD, Illumos and probably others support larger page sizes\nusing transparent page coalescing algorithms. On my FreeBSD box\nprocstat -v often shows PostgreSQL shared buffers in \"S\"-flagged\nmemor... |
[
{
"msg_contents": "Hi all,\n\npartition_bounds_equal() accesses the last datum from the given\npartition bounds directly to compare their greatest moduli. Rest of\nthe code uses get_greatest_modulus() to get the greatest modulus from\na partition bound. partition_bounds_equal() should also do the same\nfor the ... |
[
{
"msg_contents": "src/tools/msvc/builddoc.pl and the associated documentation (in\ninstall-windows.sgml) is quite outdated, even for PG10. If someone has\nthe required knowledge, please consider supplying an update.\n\n-- \nPeter Eisentraut http://www.2ndQuadrant.com/\nPostgreSQL Development, 24x... |
[
{
"msg_contents": "The following bug has been logged on the website:\n\nBug reference: 14941\nLogged by: Lyes Ameddah\nEmail address: lyes.amd@gmail.com\nPostgreSQL version: 9.6.0\nOperating system: CentOs 7\nDescription: \n\nHello, \r\n\r\nI make a complete empty once a week in an a... |
[
{
"msg_contents": "Hi,\n\nWe recently had an issue in production, where a bitmap scan was chosen \ninstead of an index scan. Despite being 30x slower, the bitmap scan had \nabout the same cost as the index scan.\n\nI've found some cases where similar issues with bitmap scans were \nreported before:\n\nhttps://w... |
[
{
"msg_contents": "Hello,\n\nI have a typical star schema, having dimension tables \"product\", \"calendar\"\nand \"country\" and a fact table \"sales\".\nThis fact table is partitionned by time (range by month) and country (list).\n\nWill query like:\n\nselect product.name, calendar.month, sum(sales.net_price)... |
[
{
"msg_contents": "Hi\n\nI am not sure if user postgres should be removed, so it is probably bug\n\npg_dumpall -r -c | grep postgres\n\nDROP ROLE postgres;\nCREATE ROLE postgres;\n\nRegards\n\nPavel\n\nHiI am not sure if user postgres should be removed, so it is probably bugpg_dumpall -r -c | grep postgresDROP ... |
[
{
"msg_contents": "I've been working on adding a new feature to pg_hexedit [1] to allow\nit to work with GIN indexes. This led to an idea about how we could do\nbetter within nbtree, which I will now describe.\n\nI noticed that GIN's internal posting tree pages (B-Trees over TIDs)\ndo away with the need for an ... |
[
{
"msg_contents": "Hi all,\n\nOn the recent following thread problems around the use of\nAllocateDir() have been diagnosed with its use in the backend code:\nhttps://www.postgresql.org/message-id/20171127093107.1473.70477@wrigleys.postgresql.org\n\nI had a close look at all the callers of AllocateDir() and noti... |
[
{
"msg_contents": "Hi,\n\nI was looking through the archives one day for a few topics that\ninterest me, and saw there was continued interest in making bootstrap\ndata less painful [1] [2]. There were quite a few good ideas thrown\naround in those threads, but not much in the way of concrete results.\nI took a ... |
[
{
"msg_contents": "Adds the GROUPS option to the window framing clause. This further\nresolves TODO list item \"Implement full support for window framing\nclauses\" and implements SQL:2011 T620. No other mainstream db has this\nfeature.\n\nApply this on top of my previous patch available here:\nhttps://www.post... |
[
{
"msg_contents": "The recent changes in commit dc6c4c9dc caused Coverity to start\ncomplaining that RemovePgTempFilesInDir calls rmdir() without checking\nits return value, as would be good practice. Now, this wasn't really a\nfault of that commit, because the code was already ignoring the result of\nunlink()... |
[
{
"msg_contents": "I'm curious about handling signals in a background worker.\n\nAs I understand it, these are blocked when the BGW is born, until\nenabled with BackgroundWorkerUnblockSignals() after possible\ncustomization.\n\nIs there a known, default behavior that some signals will\nhave, if I simply Backgro... |
[
{
"msg_contents": "Why is it that do_pg_start_backup() expects its callers to pass in\nan open DIR pointer for the pg_tblspc directory, rather than doing\nthe AllocateDir call for itself? This complicates the callers, provides\nno flexibility (since do_pg_start_backup() is hardwired to know that\npg_tblspc is ... |
[
{
"msg_contents": "While working through Michael Paquier's patch to clean up inconsistent\nusage of AllocateDir(), I noticed that ResetUnloggedRelations and its\nsubroutines are not consistent about whether a directory open failure\nresults in erroring out or just emitting a LOG message and continuing.\nResetUn... |
[
{
"msg_contents": "When VACUUM or ANALYZE skips a concurrently dropped table, log it.\n\nHopefully, the additional logging will help avoid confusion that\ncould otherwise result.\n\nNathan Bossart, reviewed by Michael Paquier, Fabrízio Mello, and me\n\nBranch\n------\nmaster\n\nDetails\n-------\nhttps://git.pos... |
[
{
"msg_contents": "Hi,\n\nCurrently, txid_current and friends export a 64-bit format of\ntransaction id that is extended with an “epoch” counter so that it\nwill not wrap around during the life of an installation. The epoch\nvalue it uses is based on the epoch that is maintained by checkpoint\n(aka only check... |
[
{
"msg_contents": "Hi,\n\nJulien Rouhaund, who has proposed a patch for partition-wise ordering\nmentioned to me offlist that the comments for PartitionBoundInfoData\ndo not mention the fact that the datums in datums array are ordered. I\nthink that's important to mention there. So here's patch to do that.\n\nT... |
[
{
"msg_contents": "Hi all\n\nTL;DR: we should delete pg_replslot/$SLOTNAME/* at the start of each\ndecoding session or we can accidentally re-use stale reorder buffer\ncontents from prior runs and $BADNESS happens. StartupReorderBuffer() is\nnot sufficient.\n\n\nDetails:\n\nPetr and I have found a bug in logica... |
[
{
"msg_contents": "As part of PGConf.Asia 2017 in Tokyo, we had an unconference topic about\nzero-downtime upgrades. After the usual discussion of using logical\nreplication, Slony, and perhaps having the server be able to read old\nand new system catalogs, we discussed speeding up pg_upgrade.\n\nThere are clu... |
[
{
"msg_contents": "Hackers,\n\nI'm getting the error in $subject and am only posting here because (a) the comments\nin src/backend/utils/mmgr/dsa.c circa line 720 suggest that this is a bug, and (b) the\nDSA code appears to be fairly new to the postgresql project, and perhaps not fully\ndebugged.\n\nI am runnin... |
[
{
"msg_contents": "Support Parallel Append plan nodes.\n\nWhen we create an Append node, we can spread out the workers over the\nsubplans instead of piling on to each subplan one at a time, which\nshould typically be a bit more efficient, both because the startup\ncost of any plan executed entirely by one worke... |
[
{
"msg_contents": "Hi,\n\nI am trying to bundle a script file with my extension. Script has event triggers and trigger function does a select query on pg_event_trigger_ddl_commands .\n\nWhile executing \"create extension\" following error is thrown :\n\n\nERROR: pg_event_trigger_ddl_commands() can only be call... |
[
{
"msg_contents": "Attached is a patch for $subject. It's based on Serge Reilau's patch of\nabout a year ago, taking into account comments made at the time. with\nbitrot removed and other enhancements such as documentation.\n\nEssentially it stores a value in pg_attribute that is used when the\nstored tuple is ... |
[
{
"msg_contents": "Hi hackers,\n\nAs far as I remember, several years ago when implementation of \nintra-query parallelism was just started there was discussion whether to \nuse threads or leave traditional Postgres process architecture. The \ndecision was made to leave processes. So now we have bgworkers, shar... |
[
{
"msg_contents": "Dear hackers,\n\nI would be happy when checkRuleResultList in rewriteDefine.c lost so strict\nconditions for returning clause. Are there any reasons to have such\nrestriction for views ?\nWhat I mean in short:\n\n create view ... as select MAIN_TABLE ... joins ...\n\n1. Updating view means up... |
[
{
"msg_contents": "Hi,\n\nI hope this is the appropriate list to send this to.\n\n*Context:*\n\nI (grad student) am trying to insert my own cardinality estimates into the\noptimizer\n\n*What I need to do to get there:*\n\nI want to be able to programmatically access the relation names inside from\ninside the ca... |
[
{
"msg_contents": "-Hackers,\n\nIn the docs it says:\n\n\"\nIf the table does not have any suitable key, then it can be set to \nreplica identity“full”, which means the entire row becomes the key.\n\n\"\n\nHow does that work? Is it using one of the hidden columns on a row?\n\n\nThanks,\n\n\nJD\n\n\n-- \nCommand... |
[
{
"msg_contents": "Hi\n\nAttached patch adds an \"%r\" substitution for psql prompts to show\nrecovery status. Specifically it displays an \"&\" (ampersand) if\nthe server is in recovery, otherwise nothing, e.g.:\n\n postgres=&# SELECT foo;\n\nWhy is this useful? Because I find myself messing about with rep... |
[
{
"msg_contents": "Hi,\n\nI just noticed a comment which has been made a little outdated by the\npartition-wise join code from commit f49842d1. The comment claims that\ninheritance children don't add to the effort required in join\nplanning, while that still may be true, we should probably mention\nthat partiti... |
[
{
"msg_contents": "A customer recently reported a crash in a postgres backend. The backend\nencountered a SIGSEGV, crashing during SELECTs from a fairly\ncomplicated view using a grouping set directive. I've managed to\nreproduce it by tracking it down to a specific SELECT, but\nunfortunately couldn't yet manag... |
[
{
"msg_contents": "Hello! I hope this is the right list for extension dev questions?\n\nI am finding odd behavior running make installcheck for a postgres\nextension. The user running the suite is a superuser (\"jfinzel\"). Once I\ncreate any object as a test role, reset role does not work, although it\ndoes... |
[
{
"msg_contents": "Some of us unfortunately have to work with multiple databases like Oracle\nor MySQL. Their respective clients mysql and sqlplus uses \"quit\" or \"exit\"\nto exit sql client.\n\nOracle's sqlplus uses \"quit\" or \"exit\" and MySQL client can be exited using\n\"quit\" and \"exit\" but for comp... |
[
{
"msg_contents": "Hi,\n\nI have a set of tables with fairly large number of columns, mostly int with\na few bigints and short char/varchar columns. I¹ve noticed that Postgres is\npretty slow at inserting data in such a table. I tried to tune every\npossible setting: using unlogged tables, increased shared_buff... |
[
{
"msg_contents": "I've been fooling around with rewriting plpgsql's composite-variable\nhandling, along the way to getting it to handle domains over composite.\nI noticed that there's some really unnecessary complication in places\nwhere it insists on separating \"row\" variables from \"record\" variables.\nMo... |
[
{
"msg_contents": "Hello,\n\nWhen I run a select query, e.g. select id from t, all columns in table \"t\" are checked to see if a column named \"id\" exists or not, and a Var is created for \"id\" if the column does exist.\n\nFunction scanRTEForColumn() does this job.\n\nBut I see in scanRTEForColumn(), the lo... |
[
{
"msg_contents": "Hi.\n\nI wonder if ScalarArrayOpExpr is not really meant for multi-dimensional\narrays appearing on the right hand side? Because:\n\n# select array[1] = any (array[array[1], array[2]]);\n\nERROR: operator does not exist: integer[] = integer\nLINE 1: select array[1] = any (array[array[1], ar... |
[
{
"msg_contents": "generate_partition_wise_join_paths prints debug info, if\nOPTIMIZER_DEBUG, using debug_print_rel at the end of each iteration for\ncollecting non-dummy child-joins, but I noticed that we pass to that\nfunction the parent's RelOptInfo, not such a child-join's RelOptInfo. I\ndon't think it's i... |
[
{
"msg_contents": "Hi.\n\nI noticed that if you partition using a array type column, partition\npruning using constraint exclusion fails to work due to a minor problem.\n\nExample:\n\ncreate table p (a int[]) partition by list (a);\ncreate table p1 partition of p for values in ('{1}');\ncreate table p1 partitio... |
[
{
"msg_contents": "Prohibit identity columns on typed tables and partitions\n\nThose cases currently crash and supporting them is more work then\noriginally thought, so we'll just prohibit these scenarios for now.\n\nAuthor: Michael Paquier <michael.paquier@gmail.com>\nReviewed-by: Amit Langote <Langote_Amit_f8... |
[
{
"msg_contents": "Hi,\n\nLooks like a minor typo in the recent commit.\n\ns/identify/identity/\n\nhttps://git.postgresql.org/gitweb/?p=postgresql.git;a=commitdiff;h=a2c6cf36608e10aa223fef49323b5feba344bfcf\n\n-\nrobins | mobile\n\nHi,Looks like a minor typo in the recent commit.s/identify/identity/https://git.... |
[
{
"msg_contents": "While reviewing the parallel CREATE INDEX patch, I noticed that commit\n8b304b8b omitted to remove a comment that it made obsolete.\n\nAttached patch removes the comment.\n\n-- \nPeter Geoghegan",
"msg_date": "Fri, 8 Dec 2017 15:41:44 -0800",
"msg_from": "Peter Geoghegan <pg@bowt.ie>"... |
[
{
"msg_contents": "Consider\n\nregression=# do $$\nregression$# declare r record;\nregression$# begin\nregression$# raise notice '%', r;\nregression$# end$$;\nERROR: record \"r\" is not assigned yet\nDETAIL: The tuple structure of a not-yet-assigned record is indeterminate.\nCONTEXT: SQL statement \"SELECT... |
[
{
"msg_contents": "The *PostgreSQL Community Weight Loss Challenge*\n\nAt PGConf Austin last week, a number of us made an early New Year's\nresolution to shed some unwanted pounds before the PGConf US National\nConference in New York City on April 16th 2018.\n\nWhile it is undeniably true that these resolutions... |
[
{
"msg_contents": "Two buildfarm runs[1][2] from the last 90 days have failed in\nsrc/test/authentication, like this:\n\n t/001_password.pl .. \n Failed 3/8 subtests \n t/002_saslprep.pl .. ok 1815 ms ( 0.00 usr 0.00 sys + 0.89 cusr 0.26 csys = 1.15 CPU)\n\n Test Summary Report\n ------------------... |
[
{
"msg_contents": "While fooling around with a different problem, I got annoyed at how slow\nMemoryContext creation and deletion is. A big chunk of that seems to be\nthe palloc/pfree cycle needed to allocate the context header block in\nTopMemoryContext. After thinking about it for a bit, I decided that\nit'd... |
[
{
"msg_contents": "I noticed that since I put in commit 390d58135, buildfarm members\nthat use CLOBBER_CACHE_ALWAYS are failing the added test case\nwith diffs like\n\n*** 5051,5057 ****\n NOTICE: y = 2\n ERROR: record \"r\" is not assigned yet\n DETAIL: The tuple structure of a not-yet-assigned record is... |
[
{
"msg_contents": "Hi hackers,\n\nI heard a report of a 10.1 cluster hanging with several 'BtreePage'\nwait_events showing in pg_stat_activity. The query plan involved\nParallel Index Only Scan, and the table is concurrently updated quite\nheavily. I tried and failed to make a reproducer, but from the clues\n... |
[
{
"msg_contents": "Hello,\n\nFYI\nI collected Web resources to learn PostgreSQL internals in the Developer FAQ page. I did this because I need to provide self-education materials for new developers. Specifically, I modified \"How is the source code organized?\", and added \"What information is available to le... |
[
{
"msg_contents": "I came across this paper making a case for indices that use machine\nlearning to optimise search.\n\nhttps://arxiv.org/pdf/1712.01208.pdf\n\nThe gist seems to be to use a linear regression model or feed a tensor flow\nmodel when a more complicated distribution is needed for the data and allow... |
[
{
"msg_contents": "Hi,\n\nWhile testing something , I found that even after rule has dropped not \nable to insert data and in an another scenario , there is a Crash/\n\nPlease refer this scenario's -\n\n1) Rows not inserted after dropping the RULE\n\npostgres=# create table e(n int);\nCREATE TABLE\npostgres=#... |
[
{
"msg_contents": "Hello -\n\nI understand how to setup a regression suite for a postgres extension, but\nwhat I'm not clear on from the docs is if there is a pattern that exists\nfor testing not only the latest version of an extension, but also an\nupgraded previous version.\n\nFor example, I am developing ver... |
[
{
"msg_contents": "Very interesting read: https://arxiv.org/abs/1712.01208\n\nHN discussion: https://news.ycombinator.com/item?id=15894896\n\nSome of the comments (from Twitter\nhttps://twitter.com/schrockn/status/940037656494317568): \"Jeff Dean and co\nat GOOG just released a paper showing how machine-learned... |
[
{
"msg_contents": "Hi,\n\nThe money type has overflow handling in its input function:\nDatum\ncash_in(PG_FUNCTION_ARGS)\n...\n\t\t\tCash\t\tnewvalue = (value * 10) - (*s - '0');\n\n\t\t\tif (newvalue / 10 != value)\n\t\t\t\tereport(ERROR,\n\t\t\t\t\t\t(errcode(ERRCODE_NUMERIC_VALUE_OUT_OF_RANGE),\n\t\t\t\t\t\t ... |
[
{
"msg_contents": "Hi.\n\nHoriguchi-san pointed out [1] on a nearby thread that the partitioning\nsyntax (the FOR VALUES clause) doesn't accept true and false as valid\npartition bound datums, which seems to me like an oversight. Attached a\npatch to fix that.\n\ncreate table bools (a bool) partition by list (... |
[
{
"msg_contents": "I have, more or less, this classic question:\n\nhttps://www.postgresql.org/message-id/4667C403.1070807%40t3go.de\n\nBut I have it for a newer reason, where again it seems as if a better\nanswer than \"don't do that\" might be worth having.\n\n1. Suppose you are running PG in a VM (named pgvm ... |
[
{
"msg_contents": "Hi all\n\nTL;DR: Lets add a ProcSignalReason that makes a backend\ncall MemoryContextStats when it sees it and a C func that users can use to\nset it on a proc. Sane?\n\n\nI fairly frequently want to get a memory context dump from a running\nbackend. It's trivial to do on a debug-enabled buil... |
[
{
"msg_contents": "Hi,\n\nworking on overflow correctness in pg I noticed that pgbench isn't quite\nthere. I assume it won't matter terribly often, but the way it parses\nintegers makes it incorrect for, at least, the negativemost number.\n\nIt directly parses the integer input using:\n{digit}+\t\t{\n\t\t\t\t\... |
[
{
"msg_contents": "\"The minimum across all databases, in turn, is recorded in checkpoint\nrecords, and CHECKPOINT removes pg_multixact/ segments older than that\nvalue once the checkpoint record has been flushed.\"\n\nThe above part of the paragraph in README.tuplock seems to be\noutdated. I think now we trun... |
[
{
"msg_contents": "Hi,\n\nThis is an awesome paper about a new index called \"Learned Index\".\nit's aka dense hash structure derived (\"learned\") from actual data.\nLooks very promising for certain properties [*].\n\nAnyone already working on this in Postgres?\nHow could this be implemented in Postgres?\n\n:S... |
[
{
"msg_contents": "While working on transaction control in procedures, I noticed some\ninconsistencies in how portal pinning is used.\n\nThis mechanism was introduced in\neb81b6509f4c9109ecf8839d8c482cc597270687 to prevent user code from\nclosing cursors that PL/pgSQL has created internally, mainly for FOR\nloo... |
[
{
"msg_contents": "One of our challenges we have is that our engineers have written frameworks\nto backfill data in several different DSLs, and every time they adopt a new\nlanguage, they maybe need to write another one.\n\nTo be clear, what I mean is batch updating a large set of data in small\npieces so as to... |
[
{
"msg_contents": "I'm working on improving coverage and would like to generate some\nreports (other than the text versions) to help me find uncovered code.\nHowever, my source path and build path are not the same and I'm running\ninto problems.\n\nThis works fine and produces gcov output:\n\n$ make -C test/bui... |
[
{
"msg_contents": "Including unlogged relations in base backups takes up space and is \nwasteful since they are truncated during backup recovery.\n\nThe attached patches exclude unlogged relations from base backups except \nfor the init fork, which is required to recreate the main fork during \nrecovery.\n\n* e... |
[
{
"msg_contents": "Provide overflow safe integer math inline functions.\n\nIt's not easy to get signed integer overflow checks correct and\nfast. Therefore abstract the necessary infrastructure into a common\nheader providing addition, subtraction and multiplication for 16, 32,\n64 bit signed integers.\n\nThe n... |
[
{
"msg_contents": "I noticed that the output of pg_localtime() in str_time() is not checked \nand can sometimes return NULL. It's pretty unlikely to occur, I guess if \nthe time() function was acting funny. For example if I define this:\n\ntime_t\nfake_time(void *blah)\n{\n return 0x0110000000000000;\n}\n\n... |
[
{
"msg_contents": "Hi,\n\nBuildfarm animal thrips just failed with a curious error:\nhttps://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=thrips&dt=2017-12-13%2002%3A27%3A27\n\n============== shutting down postmaster ==============\npg_ctl: could not open PID file \"C:/buildfarm/buildenv/HEAD/p... |
[
{
"msg_contents": "Hi hackers,\n\nThe start-up cost of bounded (top-N) sorts is sensitive at the small\nend of N, and the\ncomparison_cost * tuples * LOG2(2.0 * output_tuples) curve doesn't\nseem to correspond to reality. Here's a contrived example that comes\nfrom a benchmark:\n\n create table foo as\n se... |
[
{
"msg_contents": "HI,\n\nConsider the below test:\n\npostgres@110311=#call sum(2+2);\nserver closed the connection unexpectedly\n This probably means the server terminated abnormally\n before or while processing the request.\nThe connection to the server was lost. Attempting reset: Failed.\n\nCall stack:... |
[
{
"msg_contents": "Hi.\n\nI recently posted to the list about a couple of problems I saw when using\narray type column as the partition key. One of them was that the internal\npartition constraint expression that we generate for list partitions is of\na form that the backend would reject if the partition key c... |
[
{
"msg_contents": "Hi hackers,\n\nI suppose that extensions are supposed to be allowed to use the\nfacilities in access/parallel.h. I noticed in passing when I wrote a\nthrowaway test harness that my Windows built drone complained:\n\ntest_sharedtuplestore.obj : error LNK2001: unresolved external symbol\nParal... |
[
{
"msg_contents": "Hi,\n\nCurrently if some one try to drop the PROCEDURE and\nit don't have privilege or it's not an owner, than error message\nstill indicate object as FUNCTION.\n\nExample:\n\npostgres@37737=#drop procedure pro;\nERROR: must be owner of function pro\n\nThis doesn't look correct specially tha... |
[
{
"msg_contents": "Hi.\n\nYou may have guessed from $subject that the two don't work together.\n\ncreate table p (a int) partition by list (a);\ncreate table p1 partition of p for values in (1, 2) partition by list (a);\ncreate table p11 partition of p1 for values in (1);\ncreate table p12 partition of p1 for v... |
[
{
"msg_contents": "Hi,\nWe allow a function to be invoked as part of PERFORM statement in plpgsql\ndo $$\nbegin perform pg_relation_size('t1'); end; $$ language plpgsql;\nDO\n\nBut we do not allow a procedure to be invoked this way\n create procedure dummy_proc(a int) as $$\nbegin null; end;\n$$ language plpgsq... |
[
{
"msg_contents": "Well, that's annoying: buildfarm member castoroides just fell over\nwith symptoms indicating that its compiler thinks that\n__builtin_constant_p(\"string literal\") is false, thus breaking the\ncheck I installed in commit 9fa6f00b1 that AllocSetContextCreate's\nname argument is a literal.\n\n... |
[
{
"msg_contents": "Would this sample code make an even better teaching example if it\nused the existing GUC way to declare that worker_spi.naptime is\nin units of seconds?\n\nOr does it not do that for some reason I've overlooked?\n\n-Chap",
"msg_date": "Thu, 14 Dec 2017 17:41:24 -0500",
"msg_from": "Ch... |
[
{
"msg_contents": "Hi hackers,\n\nI am unable to build pgindent as it appears the pearltidy source has been\nremoved from sourceforge:\n\nsee prerequisite 2:\n\nhttps://git.postgresql.org/gitweb/?p=postgresql.git;a=blob_plain;f=src/tools/pgindent/README;hb=HEAD\n\nThanks,\nJordan Deitch\n\nHi hackers,I am unabl... |
[
{
"msg_contents": "Greetings -hackers,\n\nGoogle Summer of Code 2018 was announced back in September and they've\nchanged up the timeline a bit [1]. Specifically, they moved up the\ndates for things like the mentoring organization application deadline,\nso it's time to start working on our Idea's page for 2018... |
[
{
"msg_contents": "Hello, I noticed while an investigation that pg_prewarm is\nforgetting to register a tranche.\n\nBefore the second parameter of LWLockRegisterTranche() became\nchar * in 3761fe3, that would lead to a crash for a\n--enable-dtrace build, but currently it not likely.\n\nOn the other hand as far ... |
[
{
"msg_contents": "Hi,\n\nDebian's reproducible builds project has revealed that the full build\npath gets embedded into server/catalog/schemapg.h:\n\n/*-------------------------------------------------------------------------\n *\n * schemapg.h\n * Schema_pg_xxx macros for use by relcache.c\n *\n * Portions... |
[
{
"msg_contents": "To be able to identify more easily which package a connected server is\ncoming from, I would like to embed the (Debian) package version in the\nversion() output which is coming from PG_VERSION. It is fairly easy to\ndo that, but it requires patching configure(.in):\n\n$ ./configure VENDOR_VER... |
[
{
"msg_contents": "Hi!\n\n\nI found out the problem in exposing values of t_bits field from \nheap_page_items function.\n\nWhen the number of attributes in table is multiple of eight, t_bits \ncolumn shows double number of bits in which data fields are included.\n\nFor example:\n\n# create table tbl(f1 int, f2 ... |
[
{
"msg_contents": "Hi\n\nIt occurred to me the other day while people were talking about\npasting blocks of text creating problems, especially with tabs, that\nxterm bracketed-paste support (also works in at least putty and\nprobably others) that would block curses handling and just paste as-is\nwould be a usef... |
[
{
"msg_contents": "Hi,\n\nsqlsmith just crashed a parallel worker while testing master at\n699bf7d05c. I can reproduce it with the following recipe on a fresh\nregression database. Backtrace and query plan below as well.\n\nregards,\nAndreas\n\n--8<---------------cut here---------------start------------->8---... |
[
{
"msg_contents": "I recently fell afoul of a weird edge case while writing an extension. It seems Postgres allows for an ambiguous column reference to oid in the where clause when joining to pg_roles. It just arbitrarily chooses pg_roles.oid and ignores the conflicting name. Example:\r\n\r\n\r\npostgres=# C... |
[
{
"msg_contents": "Greetings everyone,\n\nI need a data type to represent genomic positions, which will consist of a\nstring and a pair of integers with interval logic and access methods. Sort\nof like my seg type, but more straightforward.\n\nI noticed somebody took a good care of seg while I was away for the ... |
[
{
"msg_contents": "Commit d70cf811, from 2014, promoted an Assert() within\nIndexBuildHeapScan() to a \"can't happen\" elog() error, in order to\ndetect when a parent tuple cannot be found for some heap-only tuple --\nif this happens, then it indicates corruption. I think that we should\nmake it a full ereport(... |
[
{
"msg_contents": "Hi,\n\nI was puzzled to see the following code:\n\nmy_extra->element_type = ~element_type;\n\nIt looks quite wrong, but if its right then I think it needs a comment\nto explain it. I don't see any in the area which mentions it. My best\nguess would be that it's using this to know if the type ... |
[
{
"msg_contents": "The attached fixed a small typo in json_agg_transfn.\n\n-- \n David Rowley http://www.2ndQuadrant.com/\n PostgreSQL Development, 24x7 Support, Training & Services",
"msg_date": "Mon, 18 Dec 2017 02:43:15 +1300",
"msg_from": "David Rowley <david.rowley@2ndquadrant.com... |
[
{
"msg_contents": "Hi,\n\nWhile working on partial aggregation a few years ago, I didn't really\nthink it was worthwhile allowing partial aggregation of string_agg and\narray_agg. I soon realised that I was wrong about that and allowing\nparallelisation of these aggregates still could be very useful when\nmany ... |
[
{
"msg_contents": "In this thread I would like to raise the issue of incremental backups.\nWhat I suggest in this thread, is to choose one direction, so we can\nconcentrate our community efforts.\nThere is already a number of tools, which provide incremental backup.\nAnd we can see five principle techniques the... |
[
{
"msg_contents": "\nI was doing some work over the weekend and it occurred to me that it\nwould be quite nice to have the input file name from regression tests\nset as the application name, and then use a log_line_prefix setting\nincluding %a, so that this would appear on the logs.\n\n\nMy first thought was to... |
[
{
"msg_contents": "I've been adding support for valgrind to the buildfarm client (code will\nhit the git repo shortly). Mostly the results have been pretty clean,\nbut the pltcl tests generated the attached output. Perhaps someone with\nmore valgrind-fu than I have so far would like to use the information to\ne... |
[
{
"msg_contents": "I was able to link ICU library with postgres as shared objects using the\nconfigure command:\n./configure --prefix=/leoaaryan/postgres-10 ... --with-icu\nICU_CFLAGS=\"-I/leoaaryan/postgres-10/include\"\nICU_LIBS=\"-L/leoaaryan/postgres-10/lib -licuuc -licudata -licui18n\"\n\nNow I'm trying li... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.