threads listlengths 1 2.99k |
|---|
[
{
"msg_contents": "Hi,\n\nWe are running a postgres streaming replica with\nmax_standby_streaming_delay set to 900 seconds (15min). We encountered an\nissue in our environment where we had a long running query that was running\nagainst the replica for just over 4 hours causing 4 hours of replication\nlag. Loo... |
[
{
"msg_contents": "Hello!\n\nFound that src/test/modules/test_misc/t/003_check_guc.pl will crash if an extension\nthat adds own GUCs was loaded into memory.\nSo it is now impossible to run a check-world with loaded extension libraries.\n\nReproduction:\ncd src/test/modules/test_misc\nexport EXTRA_INSTALL=\"cont... |
[
{
"msg_contents": "The upper planner was pathified many years ago now. That was a large\nchunk of work and because of that, the union planner was not properly\npathified in that effort. A small note was left in\nrecurse_set_operations() to mention about future work.\n\nYou can see this lack of pathification i... |
[
{
"msg_contents": "Hi hackers:\n\nI got a basebackup using pg_basebackup -R. After that, I created a restore\npoint named test on primary, and set recovery_target_name to test,\nrecovery_target_action to shutdown in standby datadir. I got a failure\nstartup message after 'pg_ctl start -D $standby_datadir'. I t... |
[
{
"msg_contents": "Hi all!\n\nCurrently, only unnamed prepared statements are supported by psql with the\n\\bind command and it's not possible to create or use named prepared statements\nthrough extended protocol.\n\nThis patch introduces 2 additional commands: \\parse and \\bindx.\n\\parse allows us to issue a... |
[
{
"msg_contents": "Any new patches will now need to be submitted to the January\ncommitfest. (Normally we'd give advance notice of that change, but\nwe're a bit behind this time.)\n\nI have started going through the patch entries looking for out-of-date\nstatuses and filling in missing author fields. Currently ... |
[
{
"msg_contents": "Hi,\n\nWhile working on something else, I noticed $SUBJECT: commit b7eda3e0e\nmoved XidInMVCCSnapshot() from tqual.c into snapmgr.c, but follow-up\ncommit c91560def updated this reference incorrectly:\n\n@@ -1498,7 +1498,7 @@ GetMaxSnapshotSubxidCount(void)\n * information may not be availab... |
[
{
"msg_contents": "This proposal showcases the speed-up provided to popcount feature when using AVX512 registers. The intent is to share the preliminary results with the community and get feedback for adding avx512 support for popcount. \n \nRevisiting the previous discussion/improvements around this feature, I... |
[
{
"msg_contents": "I have three suggestions on committing that I thought would be helpful\nto the hacker audience.\n\nFirst, I have been hesitant to ascribe others as patch authors if I\nheavily modified a doc patch because I didn't want them blamed for any\nmistakes I made. However, I also want to give them c... |
[
{
"msg_contents": "Hi,\n\n\nLook like the tab completion for CREATE TABLE ... AS is not proposed.\n\n\ngilles=# CREATE TABLE test\n( OF PARTITION OF\n\n The attached patch fix that and also propose the further completion \nafter the AS keyword.\n\n\ngilles=# CREATE TABLE test\n( ... |
[
{
"msg_contents": "Hello!\n\nI was looking into table access methods recently and found the\nexisting page a bit sparse. Here's a small patch adding a little more\nexample code to the table access methods page.\n\nLet me know if there's anything I can do to fix my patch up!\n\nCheers,\nPhil",
"msg_date": "T... |
[
{
"msg_contents": "\nI noticed that we occasionally use \"volatile\" to access shared memory,\nbut usually not; and I'm not clear on the rules for doing so. For\ninstance, AdvanceXLInsertBuffer() writes to XLogCtl->xlblocks[nextidx]\nthrough a volatile pointer; but then immediately writes to XLogCtl-\n>Initiali... |
[
{
"msg_contents": "Hi\n\nWe have one such problem. A table field has skewed data. Statistics:\nn_distinct | -0.4481973\nmost_common_vals | {5f006ca25b52ed78e457b150ee95a30c}\nmost_common_freqs | {0.5518474}\n\n\nData generation:\n\nCREATE TABLE s_user (\n user_id varchar(32) NOT NULL... |
[
{
"msg_contents": "I'm having errors restoring with pg_restore to v16.0, it appears to be a \nregression or bug. The same file restored to v15.4 without problem.\n\nDuring the restore:\n\n pg_restore: error: could not execute query: ERROR: type \"hash\" does not exist\n LINE 7: )::hash;\n [...]\n C... |
[
{
"msg_contents": "I've finished the first draft of 16.1 release notes; see\n\nhttps://git.postgresql.org/gitweb/?p=postgresql.git;a=commitdiff;h=1e774846b870a858f8eb88b3f512562009177f45\n\nPlease send comments/corrections by Sunday.\n\n\t\t\tregards, tom lane\n\n\n",
"msg_date": "Fri, 03 Nov 2023 17:56:42 ... |
[
{
"msg_contents": "Hello PostgreSQL Developers,\n\nI hope this message finds you well. My name is Atharva, and I'm a\nsecond-year student majoring in computer application with a strong interest\nin contributing to the PostgreSQL project. I have experience in C\nprogramming and a solid understanding of PostgreSQ... |
[
{
"msg_contents": "hi\n\nhttps://www.postgresql.org/docs/current/datatype-json.html\nTable 8.23. JSON Primitive Types and Corresponding PostgreSQL Types\n\n\"SQL NULL is a different concept\"\ncan we change to\n\"Only accept lowercase null. SQL NULL is a different concept\"\nI saw people ask similar questions o... |
[
{
"msg_contents": "I came across the following compiling warnings on GCC (Red Hat 4.8.5-44)\n4.8.5 with 'CFLAGS=-Og'\n\nbe-fsstubs.c: In function ‘be_lo_export’:\nbe-fsstubs.c:537:24: warning: ‘fd’ may be used uninitialized in this\nfunction [-Wmaybe-uninitialized]\n if (CloseTransientFile(fd) != 0)\n ... |
[
{
"msg_contents": "Analogous to 388e80132c (which was for Perl) but for Python, I propose \nadding #pragma GCC system_header to plpython.h. Without it, you get \ntons of warnings about -Wdeclaration-after-statement, starting with \nPython 3.12. (In the past, I have regularly sent feedback to Python to \nfix t... |
[
{
"msg_contents": "Create a table and a deferrable constraint trigger:\n\n CREATE TABLE tab (i integer);\n\n CREATE FUNCTION trig() RETURNS trigger\n LANGUAGE plpgsql AS\n $$BEGIN\n RAISE NOTICE 'current_user = %', current_user;\n RETURN NEW;\n END;$$;\n\n CREATE CONSTRAINT TRIGGER trig AFTER INSERT ON... |
[
{
"msg_contents": "Hi,\r\n\r\nAttached is the release announcement draft for the 2023-11-09 release \r\n(16.1 et al.).\r\n\r\nPlease review for accuracy and notable omissions. Please have all \r\nfeedback in by 2023-11-09 08:00 UTC at the latest (albeit the sooner the \r\nbetter).\r\n\r\nThanks,\r\n\r\nJonathan... |
[
{
"msg_contents": "(Unfortunately, I'm posting this too late for the November commitfest, but\nI'm hoping this will be the first in a series of proposed improvements\ninvolving SIMD instructions for v17.)\n\nPresently, we ask compilers to autovectorize checksum.c and numeric.c. The\npage checksum code actually... |
[
{
"msg_contents": "The following entries are WoA but have not had an email update since\nJuly. If there is no actionable update soon, I plan to mark these\nReturned with Feedback before end of CF:\n\npg_usleep for multisecond delays\nvector search support\nImprove \"user mapping not found\" error message\nParal... |
[
{
"msg_contents": "In pipeline mode after queuing a message to be sent we would flush the\nbuffer if the size of the buffer passed some threshold. The only message\ntype that we didn't do that for was the Flush message. This addresses\nthat oversight.\n\nI noticed this discrepancy while reviewing the\nPQsendSyn... |
[
{
"msg_contents": "Greetings,\n\nIf we use a Portal it is possible to open the portal and do a describe and\nthen Fetch N records.\n\nUsing a Cursor we open the cursor. Is there a corresponding describe and a\nway to fetch N records without getting the fields each time. Currently we\nhave to send the SQL \"fet... |
[
{
"msg_contents": "Hi, postgres hackers, I’m studying postgres buffer cache part. So I open this thread to communicate some buffer cache codes design and try to improve some tricky codes.\n\nFor Buffer Cache, we know it’s a buffer array, every bucket of this array is consist of a data page and its header which ... |
[
{
"msg_contents": "While working on the MERGE RETURNING patch, I noticed a pre-existing\nMERGE bug --- ExecMergeMatched() should not call ExecUpdateEpilogue()\nif ExecUpdateAct() indicates that it did a cross-partition update.\n\nThe immediate consequence is that it incorrectly tries (and fails) to\nfire AFTER ... |
[
{
"msg_contents": "Found this issue during my Fedora 39 upgrade. Tested that uninstalling \nopenssl still allows the various ssl tests to run and succeed.\n\n-- \nTristan Partin\nNeon (https://neon.tech)",
"msg_date": "Tue, 07 Nov 2023 16:06:56 -0600",
"msg_from": "\"Tristan Partin\" <tristan@neon.tech>... |
[
{
"msg_contents": "Hi,\n\nWhile looking at GUC messages today I noticed the following:\n\ngettext_noop(\"The server will use the fsync() system call in several\nplaces to make \"\n \"sure that updates are physically written to disk. This insures \"\n \"that a database cluster will recover to a consistent stat... |
[
{
"msg_contents": "Hello!\n\nI noticed that the same block\n\n-- SET statements.\n-- These use two different strings, still they count as one entry.\nSET work_mem = '1MB';\nSet work_mem = '1MB';\nSET work_mem = '2MB';\nRESET work_mem;\nSET enable_seqscan = off;\nSET enable_seqscan = on;\nRESET enable_seqscan;\n... |
[
{
"msg_contents": "Hi all,\nI have some questions about the implementation of vector32_is_highbit_set on arm.\nBelow is the comment and the implementation for this function.\n/*\n * Exactly like vector8_is_highbit_set except for the input type, so it\n * looks at each byte separately.\n *\n * XXX x86 uses the ... |
[
{
"msg_contents": "Dear hackers,\n\nWhile tracking a buildfarm, I found that drongo failed the test pg_upgrade/003_logical_slots [1].\nA strange point is that the test passed in the next iteration. Currently I'm not\nsure the reason, but I will keep my eye for it and will investigate if it\nhappens again.\n\nI ... |
[
{
"msg_contents": "One of the new tests in the infinite interval patch has revealed a bug\nin our 64-bit integer subtraction code. Consider the following:\n\nselect 0::int8 - '-9223372036854775808'::int8;\n\nThis should overflow, since the correct result (+9223372036854775808)\nis out of range. However, on plat... |
[
{
"msg_contents": "I found this in our logs, and reproduced it under v11-v16.\n\nCREATE TABLE t(a int, b int);\nINSERT INTO t SELECT generate_series(1,999);\nCREATE STATISTICS t_stats ON a,b FROM t;\n\nwhile :; do psql postgres -qtxc \"ANALYZE t\"; done &\nwhile :; do psql postgres -qtxc \"begin; DROP STATISTIC... |
[
{
"msg_contents": "clang and gcc both now support -fsanitize=address,undefined. These are \nreally useful to me personally when trying to debug issues. \nUnfortunately ecpg code has a ton of memory leaks, which makes builds \nreally painful. It would be great to fix all of them, but I don't have \nthe patience ... |
[
{
"msg_contents": "Hi,\n\nI happened to notice that some GUC names \"max_fsm_pages\" and\n\"max_fsm_relations\" are still mentioned in these translation files\n(from the REL_16_1 source zip)\n\nsrc\\backend\\po\\fr.po\nsrc\\backend\\po\\tr.po\n\n~~\n\nShould those be removed?\n\nThere was a commit [1] that said... |
[
{
"msg_contents": "Hi,\nI am trying to build postgres with meson on Windows. And I am stuck in\nthe process.\n\nSteps I followed:\n\n1. I clone postgres repo\n\n2.Installed meson and ninja\npip install meson ninja\n\n3. Then running following command:\nmeson setup build --buildtype debug\n\n4. Then I ran\ncd bu... |
[
{
"msg_contents": "doc: fix wording describing the checkpoint_flush_after GUC\n\nReported-by: Evan Macbeth\n\nDiscussion: https://postgr.es/m/155208475619.1380.12815553062985622271@wrigleys.postgresql.org\n\nBackpatch-through: master\n\nBranch\n------\nmaster\n\nDetails\n-------\nhttps://git.postgresql.org/pg/c... |
[
{
"msg_contents": "Hi,\n\nI just created a primary with wal_segment_size=512. Then tried to create a\nstandby via pg_basebackup. The pg_basebackup appeared to just hang, for quite\na while, but did eventually complete. Over a minute for an empty cluster, when\nusing -c fast.\n\nIn this case I had used wal_sync_... |
[
{
"msg_contents": "In MobilityDB\nhttps://github.com/MobilityDB/MobilityDB\nwe have defined a tstzspan type which is a fixed-size equivalent of the tstzrange type in PostgreSQL.\n\nWe have a span_union aggregate function which is the equivalent of the range_agg function in PostgreSQL defined as follows\n\nCREAT... |
[
{
"msg_contents": "Moving this to a new thread and adding it to the January commitfest.\n\nOn Thu, Nov 09, 2023 at 03:27:33PM -0600, Nathan Bossart wrote:\n> On Tue, Nov 07, 2023 at 04:58:16PM -0800, Andres Freund wrote:\n>> However, even if there's likely some other implied memory barrier that we\n>> could pig... |
[
{
"msg_contents": ">I wonder if it's worth providing a set of \"locked read\" functions.\n\nMost out-of-order machines include “read acquire” and “write release” which are pretty close to what you’re suggesting. With the current routines, we only have “read relaxed” and “write relaxed”. I think implementing ac... |
[
{
"msg_contents": "An \"en_US\" user doing:\n\n CREATE TABLE foo(t TEXT PRIMARY KEY);\n\nis providing no indication that they want an index tailored to their\nlocale. Yet we are creating the index with the \"en_US\" collation and\ntherefore imposing huge performance costs (something like 2X slower\nindex buil... |
[
{
"msg_contents": "Hi,\n\nI was adding support for the new pg_stat_statements JIT deform_counter in PoWA\nwhen I realized that those were added after jit_generation_time in the\ndocumentation while they're actually at the end of the view. I'm adding Daniel\nin Cc as the committer of the original patch.\n\nIt l... |
[
{
"msg_contents": "Hello,\n\nI'm not sure if these ideas were circulated before or not.\nWe use auto_explain a lot to investigate slow/problematic queries.\nOne of the main issues with its usability comes from the fact that EXPLAIN\noutput is logged rather than returned to the caller in some way. If you\nhave a... |
[
{
"msg_contents": "Hackers,\n\nAs promised in [1], attached are some basic tests for the low-level \nbackup method.\n\nThere are currently no tests for the low-level backup method. \npg_backup_start() and pg_backup_stop() are called but not exercised in \nany real fashion.\n\nThere is a lot more that can be don... |
[
{
"msg_contents": "We're about 1/3 of the way through.\n\nAt start:\nNeeds review: 210. Waiting on Author: 42. Ready for Committer: 29.\nCommitted: 55. Withdrawn: 10. Returned with Feedback: 1. Total: 347.\n\nToday:\nNeeds review: 197. Waiting on Author: 45. Ready for Committer: 27.\nCommitted: 63. Withdrawn: 1... |
[
{
"msg_contents": "Greetings,\n\nI am getting the following error\nbuilding on HEAD\n\nLibrary crypto found: YES\nChecking for function \"CRYPTO_new_ex_data\" with dependencies -lssl,\n-lcrypto: NO\n\nI have openssl 1.1.1 installed\n\nDave Cramer\n\nGreetings,I am getting the following errorbuilding on HEADLibr... |
[
{
"msg_contents": "In PostgreSQL, when a backend process crashes, it can cause other backend processes to also require a restart, primarily to ensure data consistency. I understand that the correct approach is to analyze and identify the cause of the crash and resolve it. However, it is also important to be abl... |
[
{
"msg_contents": "When creating a partitioned index, the partition key must be a subset of \nthe index's columns. DefineIndex() explains:\n\n * If this table is partitioned and we're creating a unique index, \nprimary\n * key, or exclusion constraint, make sure that the partition key is a\n * s... |
[
{
"msg_contents": "Hello Hackers!\n\nCurrently, pgbench will log individual transactions to a logfile when the\n`--log` parameter flag is provided. The logfile, however, does not include\ncolumn header. It has become a fairly standard expectation of users to have\ncolumn headers present in flat files. Without t... |
[
{
"msg_contents": "I just found myself researching the difference between MemoryContextReset()\nand MemoryContextResetAndDeleteChildren(), and it turns out that as of\ncommit eaa5808 (2015), there is none.\nMemoryContextResetAndDeleteChildren() is just a backwards compatibility\nmacro for MemoryContextReset(). ... |
[
{
"msg_contents": "Improve default and empty privilege outputs in psql.\n\nDefault privileges are represented as NULL::aclitem[] in catalog ACL\ncolumns, while revoking all privileges leaves an empty aclitem[].\nThese two cases used to produce identical output in psql meta-commands\nlike \\dp. Using something ... |
[
{
"msg_contents": "The documentation for max_connections does not mention that just by having\na higher value for max_connections, PostgreSQL will use more resources.\n\nWhile working with different customers, I noticed that several of them set\nmax_connections to very high numbers, even though they never expec... |
[
{
"msg_contents": "Hi,\n\nI noticed that in lazy_scan_heap(), when there are no indexes on the\ntable being vacuumed, we don't release the lock on the heap page buffer\nbefore vacuuming the freespace map. Other call sites of\nFreeSpaceMapVacuumRange() hold no such lock. It seems like a waste to\nhold a lock we ... |
[
{
"msg_contents": "Hi,\n\nWhen there are no indexes on the relation, we can set would-be dead\nitems LP_UNUSED and remove them during pruning. This saves us a vacuum\nWAL record, reducing WAL volume (and time spent writing and syncing\nWAL).\n\nSee this example:\n\n drop table if exists foo;\n create table fo... |
[
{
"msg_contents": "I noticed that the fallback pg_atomic_test_set_flag_impl() implementation\nthat uses atomic-exchange is giving pg_atomic_exchange_u32_impl() an extra\nargument. This appears to be copy/pasted from the atomic-compare-exchange\nversion a few lines down. It looks like it's been this way since ... |
[
{
"msg_contents": "While working on BUG #18187 [1], I noticed that we also have issues with\nhow SJE replaces join clauses involving the removed rel. As an example,\nconsider the query below, which would trigger an Assert.\n\ncreate table t (a int primary key, b int);\n\nexplain (costs off)\nselect * from t t1... |
[
{
"msg_contents": "Hi Team,\n\nGood day! I'm not able to launch Postgres PGAdmin 4 in my MAC OS, I have\ntried with both versions 15 and 16, but nothing is working. It says that it\nhas quit unexpectedly (screenshot attached). I have attached the bug report\nas well along with my system specifications below. Ki... |
[
{
"msg_contents": "I have been working on adding using thread-safe locale APIs within \nPostgres where appropriate[0]. The patch that I originally submitted \ncrashed during initdb (whoops!), so I worked on fixing the crash, which \nled me to having to touch some code in chklocale.c, which became \na frustratin... |
[
{
"msg_contents": "Hi,\n\nA while back I had proposed annotations for palloc() et al that let the\ncompiler know about which allocators pair with what freeing functions. One\nthing that allows the compiler to do is to detect use after free.\n\nOne such complaint is:\n\n../../../../../home/andres/src/postgresql/... |
[
{
"msg_contents": "Hi,\n\nAfter a recent commit 6a72c42f (a related discussion [1]) which\nremoved MemoryContextResetAndDeleteChildren(), I think there are a\ncouple of other backward compatibility macros out there that can be\nremoved. These macros are tuplestore_donestoring() which was\nintroduced by commit d... |
[
{
"msg_contents": "Hello,\n\nI work on pgrx, a Rust crate (really, a set of them) that allows\npeople to use Rust to write extensions against Postgres, exporting\nwhat Postgres sees as ordinary \"C\" dynamic libraries. Unfortunately,\nthe build system for this is a touch complicated, as it cannot simply\nrun pg... |
[
{
"msg_contents": "In the \"Allow tests to pass in OpenSSL FIPS mode\" thread [0] it was discovered\nthat 3DES is joining the ranks of NIST disallowed algorithms. The attached\npatch adds a small note to the pgcrypto documentation about deprecated uses of\nalgorithms. I've kept it to \"official\" notices such... |
[
{
"msg_contents": "Hi,\n\nI've often had to analyze what caused corruption in PG instances, where the\nsymptoms match not having had backup_label in place when bringing on the\nnode. However that's surprisingly hard - the only log messages that indicate\nuse of backup_label are at DEBUG1.\n\nGiven how crucial u... |
[
{
"msg_contents": "Hi,\n\nIt seems there's a long-standing data loss issue related to the initial\nsync of tables in the built-in logical replication (publications etc.).\nI can reproduce it fairly reliably, but I haven't figured out all the\ndetails yet and I'm a bit out of ideas, so I'm sharing what I know wi... |
[
{
"msg_contents": "Right now, if allocation fails while growing a hashtable, it's left in\nan inconsistent state and can't be used again.\n\nPatch attached.\n\n\n-- \nJeff Davis\nPostgreSQL Contributor Team - AWS",
"msg_date": "Fri, 17 Nov 2023 10:42:54 -0800",
"msg_from": "Jeff Davis <pgsql@j-davis.com... |
[
{
"msg_contents": "I had briefly experimented changing the hash table in guc.c to use\nsimplehash. It didn't offer any measurable speedup, but the API is\nslightly nicer.\n\nI thought I'd post the patch in case others thought this was a good\ndirection or nice cleanup.\n\n\n-- \nJeff Davis\nPostgreSQL Contribut... |
[
{
"msg_contents": "Hi,\n\nRight now we use PANIC for very different kinds of errors.\n\nSometimes for errors that are persistent, where crash-restarting and trying\nagain won't help:\n ereport(PANIC,\n (errmsg(\"could not locate a valid checkpoint record\")));\nor\n ereport(PANIC,\n (errmsg(... |
[
{
"msg_contents": "Hi,\n\nOn linux, many filesystems default to remounting themselves read-only when\nmetadata IO fails. I.e. one common reaction to disks failing is a previously\nread-write filesystem becoming read-only.\n\nWhen e.g. trying to create a file on such a filesystem, errno is set to\nEROFS. Writing... |
[
{
"msg_contents": "Hi,\n\nWe currently provide no way to learn about a postgres instance having\ncorruption than searching the logs for corruption events than matching by\nsqlstate, for ERRCODE_DATA_CORRUPTED and ERRCODE_INDEX_CORRUPTED.\n\nUnfortunately, there is a case of such an sqlstate that's not at all in... |
[
{
"msg_contents": "Hi hackers, \n\nI hope this message finds you well. I am reaching out to seek guidance on a specific aspect of PostgreSQL's index scanning functionality.\n\nI am currently working on a vector search extension for postgres, where I need to generate bitmaps based on filter conditions during an ... |
[
{
"msg_contents": "Hi,\n\nWhile reviewing another patch I was looking at the walsender's static\nfunction CreateReplicationSlot\n\nI found that the current logic seemed to have some unnecessary if/else\nchecking which can be simplified.\n\n~~\n\nTo summarise:\n\nCURRENT\nif (cmd->kind == REPLICATION_KIND_PHYSIC... |
[
{
"msg_contents": "Although it's not performance-critical, I think it just makes sense to break\nthe loop in replorigin_session_setup() as soon as we've found the origin.\n\n-- \nAntonin Houska\nWeb: https://www.cybertec-postgresql.com",
"msg_date": "Mon, 20 Nov 2023 10:07:53 +0100",
"msg_from": "Antoni... |
[
{
"msg_contents": "Hello\n\nLet me assume that there is a table T with columns a, b, c, d, e, f, g, and h.\n\nIf one wants to select data from all the columns except d and e, then one has to write\n\nSELECT a, b, c, f, g, h\nFROM T;\n\ninstead of writing\n\nSELECT ALL BUT (d, e)\nFROM T;\n\nor something similar... |
[
{
"msg_contents": "Hi hackers\n\nI've been looking into ways to reduce the overhead we're having in pqcomm\nand I'd like to propose a small patch to modify how socket_putmessage works.\n\nCurrently socket_putmessage copies any input data into the pqcomm send\nbuffer (PqSendBuffer) and the size of this buffer is... |
[
{
"msg_contents": "Hi,\n\nI've found a race condition in libpq. It is about the initialization of\nthe my_bio_methods static variable in fe-secure-openssl.c, which is not\nprotected by any lock. The race condition may make the initialization of\nthe connection fail, and as an additional weird consequence, it mi... |
[
{
"msg_contents": "meson: docs: Add {html,man} targets, rename install-doc-*\n\nWe have toplevel html, man targets in the autoconf build as well. It'd be odd\nto have an 'html' target but have the install target be 'install-doc-html',\nthus rename the install targets to match.\n\nReviewed-by: Christoph Berg <my... |
[
{
"msg_contents": "Patch attached.\n\nThe caller could do something similar, so this option is not necessary,\nbut it seems like it could be generally useful. It speeds things up for\nthe search_path cache (and is an alternative to another patch I have\nthat implements the same thing in the caller).\n\nThoughts... |
[
{
"msg_contents": "Dear Hackers,\n\nI would like to clarify, what the correct way is to determine that a given relation is using local buffers. Local buffers, as far as I know, are used for temporary tables in backends. There are two functions/macros (bufmgr.c): SmgrIsTemp, RelationUsesLocalBuffers. The first f... |
[
{
"msg_contents": "Hi all,\n\nWhile rebasing a patch from 2016 related to sequence AMs (more about\nthat later), I've bumped on a mistake from 8586bf7ed888 in\nopr_sanity.sql, as of:\n+SELECT p1.oid, p1.amname, p2.oid, p2.proname\n+FROM pg_am AS p1, pg_proc AS p2\n+WHERE p2.oid = p1.amhandler AND p1.amtype = 's... |
[
{
"msg_contents": "Hi,\n\nI decided to do some stress-testing of the built-in logical replication,\nas part of the sequence decoding work. And I soon ran into an undetected\ndeadlock related to ALTER SUBSCRIPTION ... REFRESH PUBLICATION :-(\n\nThe attached bash scripts triggers that in a couple seconds for me. ... |
[
{
"msg_contents": "Hi all,\n\nWas doing a relation size estimation based on pg_class.relpages of the\nrelation and the related objects (index, toast) and noticed that it is not\nupdated for the toast index, for example:\n\nfabrizio=# CREATE TABLE t(c TEXT);\nINSERT INTO t VALUES (repeat('x', (8192^2)::int));\n\... |
[
{
"msg_contents": "In commit 97550c0, I added a \"MyProcPid == getpid()\" check in the SIGTERM\nhandler for the startup process. This ensures that processes forked by\nsystem(3) (i.e., for restore_command) that have yet to install their own\nsignal handlers do not call proc_exit() upon receiving SIGTERM. With... |
[
{
"msg_contents": "Hi!\n\n\nCurrently libpq sends B(ind), D(escribe), E(execute), S(ync) when \nexecuting a prepared statement.\n\nThe response for that D message is a RowDescription, which doesn't \nchange during prepared\n\nstatement lifetime (with the attributes format being an exception, as \nthey aren't kn... |
[
{
"msg_contents": "=== Background\n\nSomething as simple as the following doesn't respond to cancellation. In\nv15+, any DROP DATABASE will hang as long as it's running:\n\n SELECT dblink_exec(\n $$dbname='$$||current_database()||$$' port=$$||current_setting('port'),\n 'SELECT pg_sleep(15)');\n\nhttps:/... |
[
{
"msg_contents": "Commit 3e51b278db leaves lc_* conf lines as commented-out when\ntheir value is \"C\". This leads to the following behavior.\n\n$ echo LANG\nja_JP.UTF8\n$ initdb --no-locale hoge\n$ grep lc_ hoge/postgresql.conf\n#lc_messages = 'C' # locale for system error message\n#lc_mo... |
[
{
"msg_contents": "Hi,\n\nWhile looking at the use of session_replication_state, I noticed that\nReplicationOriginLock is acquired in ReplicationOriginExitCleanup()\neven if session_replication_state is reset to NULL by\nreplorigin_session_reset(). Why can't there be a lockless exit path\nsomething like [1] sim... |
[
{
"msg_contents": "We're about 2/3 of the way through.\n\nAt start:\nNeeds review: 210. Waiting on Author: 42. Ready for Committer: 29.\nCommitted: 55. Withdrawn: 10. Returned with Feedback: 1. Total: 347.\n\nToday:\nNeeds review: 183. Waiting on Author: 45. Ready for Committer: 25.\nCommitted: 76. Returned wit... |
[
{
"msg_contents": "Hi, hackers\n\nWhen the scram_iterations value is set too large, the backend would hang for\na long time. And we can't use Ctrl+C to cancel this query, cause the loop don't\nprocess signal interrupts.\n\nAdd CHECK_FOR_INTERRUPTS within the loop of scram_SaltedPassword\nto handle any signals ... |
[
{
"msg_contents": "\nI've been trying to get meson working with Cygwin. On a fresh install \n(Cygwin 3.4.9, gcc 11.4.0, meson 1.0.2, ninja 1.11.1) I get a bunch of \nerrors like this:\n\nERROR: incompatible library \n\"/home/andrew/bf/buildroot/HEAD/pgsql.build/tmp_install/home/andrew/bf/buildroot/HEAD/inst/li... |
[
{
"msg_contents": "Hi!\n\nRecently, one of our customers had reported a not working autovacuum.\nAfter a minor investigation, I've found that\nautovacuum launcher did, actually, run vacuum as expected, but with no\nresults. At the same time, no warnings or\nother anomies were present in the logs.\n\nAt first, ... |
[
{
"msg_contents": "hi,\n\n\ni found a discuss about parallel dml, it wrote as follow,\nPostgreSQL: Re: Write operations in parallel mode\n\n\nMake updates and deletes parallel-restricted rather than\nparallel-unsafe - i.e. allow updates and deletes but only in the\nleader. This similarly would allow Update -> G... |
[
{
"msg_contents": "This patch set applies the explicit catalog representation of not-null \nconstraints introduced by b0e96f3119 for table constraints also to \ndomain not-null constraints.\n\nSince there is no inheritance or primary keys etc., this is much simpler \nand just applies the existing infrastructure... |
[
{
"msg_contents": "While working on Materialize's streaming logical replication from Postgres [0],\nmy colleagues Sean Loiselle and Petros Angelatos (CC'd) discovered today what\nappears to be a correctness bug in pgoutput, introduced in v15.\n\nThe problem goes like this. A table with REPLICA IDENTITY FULL and... |
[
{
"msg_contents": "Hi hackers,\n\nI want to discuss a new feature for assigning a snowflake ID[1], which can be\ncluster-wide unique numbers. Also, Snowflake ID can be allocated standalone.\n\n# Use case\n\nA typical use case is a multi-master system constructed by logical replication.\nThis feature allows mult... |
[
{
"msg_contents": "Dear hackers,\n\nAfter discussing the issue on irc, it looks like it could be possible \nfor the planner to use a partial index matching an expression exactly to \nestimate its selectivity.\n\nHere is a simplified version (thanks ysch) of the issue I am facing:\n\nhttps://dbfiddle.uk/flPq8-pj... |
[
{
"msg_contents": "Hello PostgreSQL Hackers,\n\nI am pleased to submit a series of patches related to the Table Access\nMethod (AM) interface, which I initially announced during my talk at\nPGCon 2023 [1]. These patches are primarily designed to support the\nOrioleDB engine, but I believe they could be benefici... |
[
{
"msg_contents": "\n\n",
"msg_date": "Thu, 23 Nov 2023 22:48:45 +0800",
"msg_from": "jacktby jacktby <jacktby@gmail.com>",
"msg_from_op": true,
"msg_subject": "Does pg support to write a new buffer cache as an extension?"
}
] |
[
{
"msg_contents": "Hi,\n\nOver in [0] and [1] there are patches that touch on the topic of\n'natual ordering' index retrieval, and [2] also touches on the topic.\nFor those patches, I've been looking at how the planner and executor\nindicate to index AMs that they expects the output to be ordered, and\nhow this... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.