threads listlengths 1 2.99k |
|---|
[
{
"msg_contents": "Pavel Stehule wrote:\n> 2012/11/27 Dimitri Fontaine <dimitri@2ndquadrant.fr>:\n\n>> I would like that we have a way to refresh a Materialized View\n>> by calling a stored procedure, but I don't think it should be\n>> the main UI.\n\nI agree. I saw that Oracle uses a function for that without ... |
[
{
"msg_contents": "Dimitri Fontaine wrote:\n> \"Kevin Grittner\" <kgrittn@mail.com> writes:\n>> An ALTER MATERIALIZED VIEW option was my first thought on syntax\n>> to do what LOAD does in the current patch. But it bothered me\n>> that I couldn't think of any other cases where ALTER\n>> <some-object-type> only ... |
[
{
"msg_contents": "Doing PITR in 9.2.1, the system claims that it reached a consistent\nrecovery state immediately after redo starts.\nThis leads to it various mysterious failures, when it should instead\nthrow a \"requested recovery stop point is before consistent recovery\npoint\" error.\n(If you are unlucky,... |
[
{
"msg_contents": "pg_basebackup is taking backup of extra files other than database related\nfiles in side a TABLESPACE directory. \n\nScenario: \n 1) Create tablespace in existing directory '/opt/tblspc' having\nsome extra files and folders. \n create tablespace tbs1 location '/opt/tbls... |
[
{
"msg_contents": "On Wed, Nov 28, 2012 at 3:55 PM, Hari Babu <haribabu.kommi@huawei.com>wrote:\n\n> pg_basebackup is taking backup of extra files other than database related\n> files in side a TABLESPACE directory.\n>\nAnd why do you have files not related to your PG server inside a folder\ndedicated to a PG s... |
[
{
"msg_contents": "A friend reported this issue to me and I find it a bit strange and even\nafter spending some time on this, I couldn't really figure out what's going\nwrong. See attached two SQL files, bad.sql and good.sql. They look the\nexact same in the editor. In fact, the good.sql is created by copying l... |
[
{
"msg_contents": "Hello Álvaro,\n\nfirst of all, thank you for bringing this up again and providing a\npatch. My first attempt on that was more than two years ago [1]. As the\nauthor of a former bgworker patch, I'd like to provide an additional\nreview - KaiGai was simply faster to sing up as a reviewer on the... |
[
{
"msg_contents": "I noticed yesterday while doing some buildfarm tinkering that the test \nscript for pg_upgrade doesn't unset various important environment values \nas pg_regress does. This can cause it to fail e.g. is PGUSER is set, as \nit was with my testing.\n\nI propose to add this to the script unless t... |
[
{
"msg_contents": "\nThis is a proposal to create some basic functions to extract values from \njson. The simple functions I envision would be:\n\n * json_object_keys(json) => setof text\n returns the set of dequoted, unescaped keys of the object,\n errors if it's not an object\n * json_get(json, keytex... |
[
{
"msg_contents": "Just a side note that the above combination doesn't work, at least not\nif the object access hook tries to make any database state updates.\nI've put a hack into index_drop that should detect the case:\n\n /*\n * We must commit our transaction in order to make the first pg_inde... |
[
{
"msg_contents": "Robert Haas wrote:\n\n> I don't particularly like syntaxes involving DO or LOAD because\n> those words already have strong associations with completely\n> unrelated features. Now, if we don't want to do that and we don't\n> want to use ALTER for a data-modifying command either, another\n> opt... |
[
{
"msg_contents": "Kevin Grittner wrote:\n\n> I still need to review the timing calls, since I'm not familiar\n> with them so it wasn't immediately obvious to me whether they\n> were being used correctly. I have no reason to believe that they\n> aren't, but feel I should check.\n\nIt seems odd to me that assign... |
[
{
"msg_contents": "Alvaro Herrera wrote:\n\n> Here's version 24.\n\nThis no longer applies after the rmgr rm_desc patch.\n\n-Kevin\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers\n",
"msg_da... |
[
{
"msg_contents": "hi\n\nWhen i test psql under multi-lingual and different encoding environment,\nI found a crash of psql.\n\n----------------------------------------------------------------------\n$ export PGCLIENTENCODING=SJIS\n$ psql\npsql (9.2rc1)\nType \"help\" for help.\n\npostgres=# \\i sql\nCREATE DATA... |
[
{
"msg_contents": "The code that reads the WAL from the archive, from pg_xlog, and from a \nmaster server via walreceiver, is quite complicated. I'm talking about \nthe WaitForWALToBecomeAvailable() function in xlog.c. I got frustrated \nwith that while working on the \"switching timeline over streaming \nrepli... |
[
{
"msg_contents": "Hi\n\nI have bumped in the segmentation fault in ecpg when using pointer to\nnested structure variable.\n\necpg crash.pgc\nSegmentation fault: 11\n\nPlease see the attached patch for the fix and test case to reproduce the\nscenario.\n\n\nThanks\nRegards\nMuhammad Usama\n\n\n\n-- \nSent via pg... |
[
{
"msg_contents": "Hi,\n\nCurrently \"make -jN check\" fails during \"creating temporary installation\"\nwith:\nmake[1]: *** read jobs pipe: Invalid argument. Stop.\nmake[1]: *** Waiting for unfinished jobs....\nmake[2]: warning: jobserver unavailable: using -j1. Add `+' to parent make rule.\nin install.log.\... |
[
{
"msg_contents": "Kevin Grittner wrote:\n> Alvaro Herrera wrote:\n> \n>> Here's version 24.\n> \n> This no longer applies after the rmgr rm_desc patch.\n\nI took a shot at merging those so I could review the patch against\nHEAD; attached in hopes that it will be useful.\n\nHopefully this isn't too far off from... |
[
{
"msg_contents": "Hi,\n\nwhile looking at trigger.c from the POV of foreign key locks I noticed\nthat GetTupleForTrigger commentless assumes it can just look at a pages\ncontent without a\n LockBuffer(buffer, BUFFER_LOCK_SHARE);\n\nThat code path is only reached for AFTER ... FOR EACH ROW ... triggers,\nso ... |
[
{
"msg_contents": "In looking to add an fsync-only option to initdb, I found its main()\nfunction to be 743 lines long, and very hard to understand.\n\nThe attached patch moves much of that code into separate functions,\nwhich will make initdb.c easier to understand, and easier to add an\nfsync-only option. Th... |
[
{
"msg_contents": "Hi,\n\nI looked at the discussion for this patch and the patch itself. Here\nare my comments and observations about the patch.\nWhat I got from the discussion is that the patch tries to implement a\nmechanism to install extension from series of SQL scripts from\nbase/full version e.g. if a us... |
[
{
"msg_contents": "Hi,\n\nThe subject says it all.\n\nThere are workloads where its detrimental, but in general having it\ndefault to on improver experience tremendously because getting conflicts\nbecause of vacuum is rather confusing.\n\nIn the workloads where it might not be a good idea (very long queries on\... |
[
{
"msg_contents": "On Wed, Nov 28, 2012 at 1:55 AM, Hari Babu <haribabu.kommi@huawei.com> wrote:\n> I think backup should be done only files and folders present inside\n> '/opt/tblspc/PG_*' directory (TABLESPACE_VERSION_DIRECTORY).\n> Not all the files and folders in side '/opt/tblspc.' directory.\n\nI think I ... |
[
{
"msg_contents": "Claudio Freire wrote:\n\n> Without hot standby feedback, reporting queries are impossible.\n> I've experienced it. Cancellations make it impossible to finish\n> any decently complex reporting query.\n\nWith what setting of max_standby_streaming_delay? I would rather\ndefault that to -1 than d... |
[
{
"msg_contents": "Claudio Freire wrote:\n\n>> With what setting of max_standby_streaming_delay? I would rather\n>> default that to -1 than default hot_standby_feedback on. That\n>> way what you do on the standby only affects the standby.\n> \n> 1d\n\nWas there actually a transaction hanging open for an entire ... |
[
{
"msg_contents": "Some of the buildfarm members are failing the pg_upgrade regression test\nsince commit 12ee6ec71f8754ff3573711032b9b4d5a764ba84. I can duplicate\nit here, and the symptom is:\n\npg_restore: creating TYPE float8range\npg_restore: creating TYPE insenum\npg_restore: [archiver (db)] Error while ... |
[
{
"msg_contents": "Someone just reported a problem when they had created a new tablespace\ninside the old data directory. I'm sure there can be other issues\ncaused by this as well, but this is mainly a confusing scenario for\npeople now.\n\nAs there isn't (as far as I know at least) any actual *point* in\ncrea... |
[
{
"msg_contents": "It's hard to know whether your tables will be locked for long periods\nwhen implementing DDL changes.\n\nThe NOREWRITE option would cause an ERROR if the table would be\nrewritten by the command.\n\nThis would allow testing to highlight long running statements before\ncode hits production.\n\... |
[
{
"msg_contents": "I'd like to change the way we set the CONTRIB_TESTDB name for contrib \nmodules. so that each module doesn't wipe out the previous module's test \ndb. The reason is that this will let us test upgrading them using \npg_upgrade much more easily. Not testing this is a significant hole in \nthe p... |
[
{
"msg_contents": "Patch for the changes discussed in\nhttp://archives.postgresql.org/pgsql-hackers/2010-10/msg00919.php\nattached (eventually ...)\n\nIn summary: If the input file (-f) doesn't exist or the ouput or log\nfiles (-o and -l) can't be created psql exits before prompting for a\npassword.\n\nRegards,... |
[
{
"msg_contents": "Add mode where contrib installcheck runs each module in a separately named database.\n\nNormally each module is tested in aq database named contrib_regression,\nwhich is dropped and recreated at the beginhning of each pg_regress run.\nThis mode, enabled by adding USE_MODULE_DB=1 to the make c... |
[
{
"msg_contents": "hi\n\nisinf() is not build to libecpg.so if build and install postgresql by \nsource on solaris9.\n(isinf() is not contained within solaris9 system.)\n\nbash-2.05$ nm /usr/local/pgsql/lib/libecpg.so | grep isinf\n[215] | 0| 0|FUNC |GLOB |0 |UNDEF |isinf\n\nIt(isinf missing... |
[
{
"msg_contents": "I was looking at the code in lazy_scan_heap() and I realized there are\ncouple of low-hanging optimizations that we can do there.\n\n1. The for-loop walks through each block of the relation. But if scan_all\nis set to false, which would be the case most often, we can jump over to\nthe next no... |
[
{
"msg_contents": ">I think we should expect provided path to be relative to current directory\n> or may consider it to be relative to either one of Data or CWD.\n>Because normally we expect path to be relative to CWD if some program is\n> asking for path in command line.\n\nPlease find the attached patch to ma... |
[
{
"msg_contents": "Marko Tiikkaja wrote:\n> Kevin Grittner <kgrittn@mail.com> wrote:\n>> Marko Tiikkaja wrote:\n>>> <T2 sees an empty table>\n>>\n>> As far as I know you are the first to notice this behavior.\n>> Thanks for pointing it out.\n>>\n>> I will take a look at the issue; I don't know whether it's\n>> ... |
[
{
"msg_contents": "I was looking at the code when the following tiny bit caught my attention.\nIn vacuumlazy.c, we release the pin on the final VM page at line number 972.\n\n 954 if (vacrelstats->num_dead_tuples > 0)\n 955 {\n 956 /* Log cleanup info before we touch indexes */\n 957 vac... |
[
{
"msg_contents": "I wonder if we really need to make another pass over the entire visibility\nmap to count the number of all-visible pages at the end of the vacuum. The\ncode that I'm looking at is in src/backend/commands/vacuumlazy.c:\n\n 247 new_rel_allvisible = visibilitymap_count(onerel);\n 248 if ... |
[
{
"msg_contents": "Jan Wieck wrote:\n\n> Attached is a new patch that addresses most of the points raised\n> in discussion before.\n> \n> 1) Most of the configuration variables are derived from\n> deadlock_timeout now. The \"check for conflicting lock request\"\n> interval is deadlock_timeout/10, clamped to 10m... |
[
{
"msg_contents": "http://git.postgresql.org/gitweb/?p=postgresql.git;a=commitdiff;h=8de72b66a2edcf12c812de0a73bd50b6b7d81d62\n\nPart of that patch was reverted later:\n\nhttp://git.postgresql.org/gitweb/?p=postgresql.git;a=commitdiff;h=5457a130d3a66db807d1e0ee2b8e829321809b83\n\nI assume that refers to the dis... |
[
{
"msg_contents": "I recently applied the attached patch to prevent recreation of the\ncurrent database user in the dump file when in binary upgrade mode. \nThis was necessary because pg_upgrade will fail on any error from a\npg_dumpall restore.\n\nIt would be nice of we could do the same for non-binary-upgrade... |
[
{
"msg_contents": "Hi all,\n\nWhile reading some index-related code, I found that the description of the\nargument is_internal of index_create in index.c has been forgotten in\ncommit f4c4335.\nCorrection patch attached.\n\nThanks,\n-- \nMichael Paquier\nhttp://michael.otacoo.com\n\n\n-- \nSent via pgsql-hacker... |
[
{
"msg_contents": "Why was this change made?\n\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers\n",
"msg_date": "Mon, 03 Dec 2012 23:24:08 -0500",
"msg_from": "Peter Eisentraut <peter_e@g... |
[
{
"msg_contents": "Hi all\n\nWhile doing some setup I've noticed that master (at least) appears to\nfail to build if LDAP headers aren't installed.\n\ngcc -O2 -Wall -Wmissing-prototypes -Wpointer-arith -Wdeclaration-after-statement -Wendif-labels -Wmissing-format-attribute -Wformat-security -fno-strict-aliasing... |
[
{
"msg_contents": "I was trying some simple queries on a Hot Standby with streaming\nreplication.\n\nOn standby, I do this:\npostgres=# begin transaction isolation level repeatable read;\nBEGIN\npostgres=# explain verbose select count(b) from test WHERE a > 100000;\n\nOn master, I insert a bunch of tuples in th... |
[
{
"msg_contents": "Hello all,\n\nthe extension modules (TestUpgrade etc.) in the buildfarm client do not \nuse the \"make\" command override defined in the config file, instead \nhardcoding command lines using \"make\". This fails where make is not GNU \nmake.\n\nPatch attached, which fixes the problem on jagua... |
[
{
"msg_contents": "Jan Wieck wrote:\n\n> Thinking about it, I'm not really happy with removing the \n> autovacuum_truncate_lock_check GUC at all.\n> \n> Fact is that the deadlock detection code and the configuration\n> parameter for it should IMHO have nothing to do with all this in\n> the first place. A proper... |
[
{
"msg_contents": "I was looking at the following code in heapam.c:\n\n 261 /*\n 262 * If the all-visible flag indicates that all tuples on the page\nare\n 263 * visible to everyone, we can skip the per-tuple visibility\ntests. But\n 264 * not in hot standby mode. A tuple that's already visib... |
[
{
"msg_contents": "Hi guys (and girls)\n\nI've been banging my head over this for a few days now so if any of you kind\nsouls could take a minute to take a look at this I would be eternally\ngrateful.\n\nI have a pretty straightforward query that is very slow by default, and\nabout 70 times faster when I set en... |
[
{
"msg_contents": "Hi everyone,\n\nAt\nhttp://git.postgresql.org/gitweb/?p=users/andresfreund/postgres.git;a=shortlog;h=refs/heads/xlogreader_v3\ngit://git.postgresql.org/git/users/andresfreund/postgres.git\nyou can find my attempt trying to bring the xlogreader from Heikki, as\nmodified by Alvaro, into a state... |
[
{
"msg_contents": "Jan Wieck wrote:\n\n> [arguments for GUCs]\n\nThis is getting confusing. I thought I had already conceded the\ncase for autovacuum_truncate_lock_try, and you appeared to spend\nmost of your post arguing for it anyway. I think. It's a little\nhard to tell. Perhaps the best thing is to present ... |
[
{
"msg_contents": "Patch looks good and also like it will/can be ready for 9.3. I'm happy\nto put time into this as committer and/or reviewer and take further\nresponsibility for it, unless others wish to.\n\nLIKES\n\n* It's pretty simple to understand and use\n\n* Check qual is stored in pre-parsed form. That ... |
[
{
"msg_contents": "Hello everybody,\n\nThere is message from Tom Lane (2008 !) about ramdisk:\n\n<begin />\n\n\n> So, IMHO, saying \"trust your OS + PostgreSQL\" is not a 100% perfect\n> approach for the people who are asking to keep their objects on RAM,\n> even though I know that there is nothing we can say r... |
[
{
"msg_contents": "I have done a high level review of the patch, and here are my comments:\n\n1) I don't know if community is really looking for this functionality, I\nfollowed the thread a bit and appraently there is a disagreement over it. I\nwill not comments on that. Let the big guys decide if they really w... |
[
{
"msg_contents": "Robert Haas wrote:\n\n> Since people *already* raise deadlock_timeout to obscenely high\n> values (a minute? an hour???) and then complain that things blow\n> up in their face, I think there's a decent argument to be made\n> that piggybacking anything else on that setting is unwise.\n\nIf peo... |
[
{
"msg_contents": "Hi all!\n\nhttp://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=guaibasaurus&dt=2012-12-05%2016%3A17%3A01\n\n\nseems like a rather odd failure in the isolation test (client)\n\n\nStefan\n\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscr... |
[
{
"msg_contents": "Stefan Kaltenbrunner wrote:\n\n> Subject: [HACKERS] strange isolation test buildfarm failure on guaibasaurus\n\n> http://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=guaibasaurus&dt=2012-12-05%2016%3A17%3A01\n\n> seems like a rather odd failure in the isolation test (client)\n\nLines which migh... |
[
{
"msg_contents": "Some idle thoughts has provoked me to sending this email. I'm wondering what\nI'm seeing here, is a simple missed optimisation or something that would be\nvery difficult to implement, or even perhaps something I've completely\nmisunderstood.\n\n \n\nTo set the scene let's go with the old prod... |
[
{
"msg_contents": "Pg_upgrade displays file names during copy and database names during\ndump/restore. Andrew Dunstan identified three bugs:\n\n* long file names were being truncated to 60 _leading_ characters, which\noften do not change for long file names\n\n* file names were truncated to 60 characters in ... |
[
{
"msg_contents": "Hello, I have a problem with PostgreSQL 9.2 with Pacemaker.\n\nHA standby sometime failes to start under normal operation.\n\nTesting with a bare replication pair showed that the standby\nfailes startup recovery under the operation sequence shown\nbelow. 9.3dev too, but 9.1 does not have this... |
[
{
"msg_contents": "In a BEFORE UPDATE trigger I need to know whether the row was previously \nmodified by this transaction. Is it safe to use xmin and txid_current() \nfor this purpose (xmin is 32-bit txid type but txid_current() returns \n64-bit bigint).\n\nCREATE FUNCTION test_trigger()\nRETURNS TRIGGER AS $$... |
[
{
"msg_contents": "Hi All,\n\nI briefly mentioned this idea in one of the other thread, but starting\na new thread to highlight the point. Today, we set the visibility map\n*only* in the first phase of vacuum. This works when the page has no\ndead tuples. But the vacuum itself is removing one or more dead tuple... |
[
{
"msg_contents": "Hello,\n\nAlthough this may have to be posted to pgsql-bugs or pgsql-general, let me \nask you here because the problem probably needs PostgreSQL's code fix.\n\n\n[Problem]\nI'm using PostgreSQL 9.1.6 on Linux. I encountered a serious problem that \nmedia recovery failed showing the followin... |
[
{
"msg_contents": "David Rowley wrote:\n\n> If we wanted to see the sales per product we could write\n> something like this:\n\n> SELECT p.product_code,SUM(s.quantity)\n> FROM products p\n> INNER JOIN bigsalestable s ON p.productid = s.productid\n> GROUP BY p.product_code;\n\n> Though this plan might not be qui... |
[
{
"msg_contents": "Robert Haas wrote:\n> Jeff Davis <pgsql@j-davis.com> wrote:\n>> Or, I could write up a test framework in ruby or python, using\n>> the appropriate pg driver, and some not-so-portable shell\n>> commands to start and stop the server. Then, I can publish that\n>> on this list, and that would at ... |
[
{
"msg_contents": "MauMau wrote:\n\n> [Problem]\n> I'm using PostgreSQL 9.1.6 on Linux. I encountered a serious\n> problem that media recovery failed showing the following message:\n> \n> FATAL: archive file \"000000010000008000000028\" has wrong size:\n> 7340032 instead of 16777216\n> \n> I'm using normal cp c... |
[
{
"msg_contents": "Tom Lane wrote:\n\n> In the case being presented here, it's not apparent to me that\n> there's any advantage to be had at all.\n\nThe OP reported a different plan which was twice as fast, although\nshowing EXPLAIN ANALYZE results for both would be nice confirmation\nof that.\n\n> You still ne... |
[
{
"msg_contents": "Background worker processes\n\nBackground workers are postmaster subprocesses that run arbitrary\nuser-specified code. They can request shared memory access as well as\nbackend database connections; or they can just use plain libpq frontend\ndatabase connections.\n\nModules listed in shared_... |
[
{
"msg_contents": "On a new buildfarm member friarbird \n<http://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=friarbird&dt=2012-12-06%2020%3A55%3A31>, \nconfigured with _DCLOBBER_CACHE_ALWAYS:\n\n BEGIN;\n TRUNCATE vistest;\n COPY vistest FROM stdin CSV FREEZE;\n + NOTICE: FREEZE option spec... |
[
{
"msg_contents": "I got a report today on the IRC channel about a pg_upgrade problem with\nupgrading clusters with indexes that exist but are invalid.\n\nFor example, if you use CREATE INDEX CONCURRENTLY, then shut down the\nserver while it is running, the index will be left as INVALID; from our\nCREATE INDEX... |
[
{
"msg_contents": "We're entering the last planned week of the CF. Apart from a 72-hour period\nin mid-November, some CF has remained in-progress continuously for the last\n176 days. With 64 out of the 82 current patches unresolved, we're on track to\nagain see no gap between CF 2012-11 and CF 2013-01. The p... |
[
{
"msg_contents": "Hello all,\n\nI have some needs that seem to support changing Postgres slightly to\ngive user programs a lot more power over how to process logging output\nthat neither the log collector nor the syslog output can well-satisfy\nas-is.\n\nI am approaching this from the angle of increasing power... |
[
{
"msg_contents": "PGCon 2013 will be on 23-24 May 2013 at University of Ottawa.\n\nThis year, we are planning to have an un-conference day around PGCon.\nThis is currently being scheduled. More information on the\nun-conference will be available within a few weeks.\n\nNOTE: the un-conference day content will ... |
[
{
"msg_contents": "Jan Wieck wrote:\n\n> Based on the discussion and what I feel is a consensus I have\n> created an updated patch that has no GUC at all. The hard coded\n> parameters in include/postmaster/autovacuum.h are\n> \n> AUTOVACUUM_TRUNCATE_LOCK_CHECK_INTERVAL 20 /* ms */\n> AUTOVACUUM_TRUNCATE_LOCK_... |
[
{
"msg_contents": "Hi Karl,\n\nI have given the patch a quick review and read the related mails\nfollowing its initial submission.\n\nI agree with that functionality along these lines is desirable. The\nability to manage output from within psql at least as richly as is\npossible with shell redirection - and cha... |
[
{
"msg_contents": "On Thu, 6 Dec 2012 10:12:31 +0530 Amit Kapila wrote:\nOn Tuesday, December 04, 2012 8:37 AM Amit Kapila wrote:\n> On Monday, December 03, 2012 8:59 PM Tom Lane wrote:\n> > Robert Haas <robertmhaas(at)gmail(dot)com> writes:\n> > > On Sat, Dec 1, 2012 at 11:45 AM, Tom Lane <tgl(at)sss(dot)pgh(d... |
[
{
"msg_contents": "Hi all,\nThe usage of the pg_database_size function generate a lot of \"permission\ndenied\" log errors.\nTypical message is: ERROR: could not stat file \"base/16384/20041\":\nPermission denied.\n\nAs a short description:\n- PostgreSQL 9.2.1/64 bits\n- our system is running on windows platfo... |
[
{
"msg_contents": "Assumption: I have enough memory to cache all the database pages.\nGoal:\nMaster never write pages. Slave replays logs from master and writes pages.\nBenefits:\nReduce the page IO overhead at master, save money in EC2 cloud.\n\nQuestion:\nCan you give me some comments on this idea?\nAnd I can... |
[
{
"msg_contents": "Andres Freund's xlogreader patch contains a change to move the \ndeclarations of SQL-callable functions in xlogfuncs.c to a new header \nfile, xlog_fn.h. The purpose is to allow xlog_internal.h to be included \nin a frontend program, as the PG_FUNCTION_ARGS and Datum used in the \nprototypes ... |
[
{
"msg_contents": "Hi,\n\nThe top-level README in the source tarball says:\n\n------------------<snip>---------------------------\nSee the file INSTALL for instructions on how to build and install\nPostgreSQL. That file also lists supported operating systems and\nhardware platforms and contains information reg... |
[
{
"msg_contents": "Hi!\n\nIt seems that PostgreSQL 9.2.2 has a regression in pg_upgrade, the -o\nand -O options forget to add a space before passing on user options,\nthereby generating unparsable command lines.\n\nFor example:\npg_upgrade -b /usr/local/pg91/bin -B /usr/bin -d /tmp/91 -D /tmp/92 -O -F\n[...]\nC... |
[
{
"msg_contents": "The documentation for PQclear() doesn't say whether it is safe to call\nPQclear() more than once on the same PGresult pointer. In fact, it is\nnot safe, but apparently only because of this last step:\n /* Free the PGresult structure itself */\n free(res);\n\nThe other members of PGresul... |
[
{
"msg_contents": "Hello\n\nSometimes I had a problems with identification custom exceptions in\nplpgsql, because we skip collecting context\n\n/*\n * error context callback to let us supply a call-stack traceback\n */\nstatic void\nplpgsql_exec_error_callback(void *arg)\n{\n PLpgSQL_execstate *estate = ... |
[
{
"msg_contents": "Hi,\n\nwith 9.2.2 I can see harmless cosmetic defect while dumping/restoring\ndatabase with postgis extension.\n\nSteps to reproduce:\n\n- create new database;\n- CREATE EXTENSION IF NOT EXISTS postgis WITH SCHEMA public;\n- backup it;\n- create new database and restore it from this new backu... |
[
{
"msg_contents": "Jan Wieck wrote:\n\n> Cleaned up all of those.\n\nApplied with trivial editing, mostly from a pgindent run against\nmodified files.\n\n-Kevin\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref... |
[
{
"msg_contents": "Hi,\n\nPlease find attached v3 of the patch to add information to event\ntriggers, including command string. This work has mostly been part of\nthe original goal and its design is the result of much reviewing here,\nin particular from Robert and Tom (rewriting a command before is done is\ncal... |
[
{
"msg_contents": "Hi,\n\nWhen initiating a new logical replication \"slot\" I want to provide a\n'SET TRANSACTION SNAPSHOT'able snapshot which can be used to setup a new\nreplica. I have some questions arround this where I could use some input\non.\n\nFirst, some basics around how this currently works:\n\nTest... |
[
{
"msg_contents": "\nI'm sure I've had a stroke or something in the middle of the night and just\ndidn't notice, but I'm able to reproduce the following on three different\nhosts on both 9.2.1 and 9.2.2. As far as I know the only difference between\nthese queries is whitespace since I just up-arrowed them in ps... |
[
{
"msg_contents": "Hi all\n\nThere's an issue with MSVC builds on Windows where clean.bat deletes\nsrc\\include\\utils\\fmgroids.h (as it should) but build.pl doesn't\nre-create it reliably.\n\nIt's created fine on the first build because Gen_fmgrtab.pl is called if\nsrc\\backend\\utils\\fmgrtab.c is missing, w... |
[
{
"msg_contents": "COPY IN loops in heap_multi_insert() extending the table until it fills the\ndisk when trying to insert a wide row into a table with a low fill-factor.\nInternally fill-factor is implemented by reserving some space space on a\npage. For large enough rows and small enough fill-factor bulk_mult... |
[
{
"msg_contents": "Currently, ANALYZE collects data on all columns and stores these\nsamples in pg_statistic where they can be seen via the view pg_stats.\n\nIn some cases we have data that is private and we do not wish others\nto see it, such as patient names. This becomes more important when we\nhave row secu... |
[
{
"msg_contents": "\nA client is testing a migration from 9.1 to 9.2, and has found that a \nlarge number of queries run much faster if they use index-only scans. \nHowever, the only way he has found to get such a plan is by increasing \nthe seq_page_cost to insanely high levels (3.5). Is there any approved \nw... |
[
{
"msg_contents": "Hi,\n\nI noticed a \"XXX: It might be worth considering using an atomic fetch-and-add\ninstruction here, on architectures where that is supported.\" in lock.c\n\nHere is my first try at using it. The patch adds a configure check for\ngcc 4.7 __atomic_add_fetch as well as the older __sync_add_... |
[
{
"msg_contents": "\nHey libpq-fans,\n\ni have question compiling libpq on windows (VS2010 32 and 64bit) with SSL support.\nI downloaded the latest source of \npostgres http://www.postgresql.org/ftp/source/v9.2.2/ and also the OpenSSL Win64 v1.0.1c http://slproweb.com/products/Win32OpenSSL.html .\nI ran nmake i... |
[
{
"msg_contents": "\r\nHi,pgsql-hackers,\r\n\r\nI'm not sure whether it is a bug of using spgist index or not .\r\n\r\nOS Version:\r\nCentOS release 6.2 (Final)\r\n\r\nPostgreSQL Version:\r\npostgres=# select version();\r\nversion\r\n------------------------------------------------------------------------------... |
[
{
"msg_contents": "As I'm working through the parallel dump patch, I notice this in one of \nthe header files:\n\n#ifdef USE_ASSERT_CHECKING\n#define Assert(condition) \\\n if (!(condition)) \\\n { \\\n write_msg(NULL, \"Failed assertion in %s, line %d\\n\", \\\n __FILE__, __LI... |
[
{
"msg_contents": "In <3688.1355509732@sss.pgh.pa.us> I complained\n> PS: one odd thing here is that the ereport(LOG) in\n> InstallXLogFileSegment isn't doing anything; otherwise we'd have gotten\n> a much more helpful error report about \"could not link file\". I don't\n> think we run the bootstrap mode with ... |
[
{
"msg_contents": "In a fully-built source tree:\n\n$ cd pgsql/src/backend/parser\n$ make\nmake: Nothing to be done for `all'.\n... okay so far ...\n$ rm gram.o\nrm: remove regular file `gram.o'? y\n$ make\nmake: Nothing to be done for `all'.\n\nWTF?\n\nIf I also remove objfiles.txt then make wakes up and remem... |
[
{
"msg_contents": "Tom Lane wrote:\n\n> the parser tables are basically number-of-tokens wide by\n> number-of-states high. (In HEAD there are 433 tokens known to the\n> grammar, all but 30 of which are keywords, and 4367 states.)\n> \n> Splitting the grammar into multiple grammars is unlikely to do\n> much to i... |
[
{
"msg_contents": "In the plperl and plpython makefiles we look for a shared library of\nlibperl or libpython, and if it's not found, we check for\nallow_nonpic_in_shlib, and if that's yes, then we proceed anyway.\nApparently, and IIRC, this was set up in a time when those shared\nlibraries were not available t... |
[
{
"msg_contents": "from 9.3devel (this morning):\n\nThere is a small bug in pg_basebackup: it displays a truncated part of what seems to be the new\n$PGDATA (the source code has 'filename' there, but I don't see what's supposed to be in there).\n\nThe truncated name in parentheses only shows up during the filli... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.