Re: Thoughts about NUM_BUFFER_PARTITIONS

> On Feb 10, 2024, at 20:15, Tomas Vondra wrote: > > On 2/8/24 14:27, wenhui qiu wrote: >> Hi Heikki Linnakangas >>I think the larger shared buffer higher the probability of multiple >> backend processes accessing the same bucket slot BufMappingLock >> simultaneously, （

Re: Transaction timeout

> 在 2023年12月23日，11:35，Junwang Zhao 写道： > > On Sat, Dec 23, 2023 at 11:17 AM Japin Li wrote: >> >> a >>> On Sat, 23 Dec 2023 at 10:40, Japin Li wrote: >>> On Sat, 23 Dec 2023 at 08:32, Japin Li wrote: On Fri, 22 Dec 2023 at 23:30, Junwang Zhao wrote: > On Fri, Dec 22, 2023 at

Re: redundant check of msg in does_not_exist_skipping

> On Nov 18, 2022, at 4:04 AM, Robert Haas wrote: > > On Thu, Nov 17, 2022 at 10:56 AM Tom Lane wrote: >> This is a completely bad idea. If it takes that level of analysis >> to see that msg can't be null, we should leave the test in place. >> Any future modification of either this code or

Re: Built-in connection pooler

On Mar 22, 2021, at 4:59 AM, Zhihong Yu mailto:z...@yugabyte.com>> wrote: Hi, + With load-balancing policy postmaster choose proxy with lowest load average. + Load average of proxy is estimated by number of clients connection assigned to this proxy with extra weight for

Re: [BUG] wrong refresh when ALTER SUBSCRIPTION ADD/DROP PUBLICATION

> On Aug 7, 2021, at 1:35 PM, Amit Kapila wrote: > > On Fri, Aug 6, 2021 at 9:57 PM Japin Li wrote: >> >>> >>> Hmm yes, it cannot cover all cases. I had somehow misunderstood that >>> the subscriber knows which relations are associated with which >>> publications. Given that the subscriber

Re: WIP: System Versioned Temporal Table

On Jan 27, 2021, at 12:39 AM, Surafel Temesgen mailto:surafel3...@gmail.com>> wrote: On Tue, Jan 26, 2021 at 2:33 PM Vik Fearing mailto:v...@postgresfriends.org>> wrote: I'm still in the weeds of reviewing this patch, but why should this fail? It should not fail. Attached is rebased patch

Re: Support ALTER SUBSCRIPTION ... ADD/DROP PUBLICATION ... syntax

> On Jan 27, 2021, at 19:41, Bharath Rupireddy > wrote: > > On Wed, Jan 27, 2021 at 4:42 PM Amit Kapila wrote: >>> On Wed, Jan 27, 2021 at 3:16 PM Bharath Rupireddy >>> wrote: On Wed, Jan 27, 2021 at 3:01 PM Amit Kapila wrote: >>> So, I think the new syntax, ALTER SUBSCRIPTION

Re: Support ALTER SUBSCRIPTION ... ADD/DROP PUBLICATION ... syntax

> On Jan 27, 2021, at 19:41, Bharath Rupireddy > wrote: > > On Wed, Jan 27, 2021 at 4:42 PM Amit Kapila wrote: >>> On Wed, Jan 27, 2021 at 3:16 PM Bharath Rupireddy >>> wrote: On Wed, Jan 27, 2021 at 3:01 PM Amit Kapila wrote: >>> So, I think the new syntax, ALTER SUBSCRIPTION

Re: Logical Replication - behavior of ALTER PUBLICATION .. DROP TABLE and ALTER SUBSCRIPTION .. REFRESH PUBLICATION

> On Jan 14, 2021, at 8:44 PM, Amit Kapila wrote: > > On Thu, Jan 14, 2021 at 6:02 PM japin wrote: >> >> On Thu, 14 Jan 2021 at 20:19, Bharath Rupireddy wrote: >>> On Thu, Jan 14, 2021 at 5:36 PM Li Japin wrote >>>> Do we really need to access PUBLIC

Re: Logical Replication - behavior of ALTER PUBLICATION .. DROP TABLE and ALTER SUBSCRIPTION .. REFRESH PUBLICATION

On Jan 14, 2021, at 1:25 PM, Amit Kapila mailto:amit.kapil...@gmail.com>> wrote: On Wed, Jan 13, 2021 at 5:40 PM Bharath Rupireddy mailto:bharath.rupireddyforpostg...@gmail.com>> wrote: On Wed, Jan 13, 2021 at 3:16 PM Bharath Rupireddy mailto:bharath.rupireddyforpostg...@gmail.com>> wrote:

Re: Fix typo about WalSndPrepareWrite

On Jan 14, 2021, at 12:56 PM, Ashutosh Bapat mailto:ashutosh.bapat@gmail.com>> wrote: Hi Japin, Thanks for the report. I think that comment is correct. It refers to the following code blocks of XLogSendPhysical() 2744 /* 2745 * OK to read and send the slice. 2746 */ 2747

Re: Logical Replication - behavior of ALTER PUBLICATION .. DROP TABLE and ALTER SUBSCRIPTION .. REFRESH PUBLICATION

On Jan 12, 2021, at 5:47 PM, japin mailto:japi...@hotmail.com>> wrote: On Tue, 12 Jan 2021 at 14:38, Amit Kapila wrote: On Tue, Jan 12, 2021 at 11:39 AM Bharath Rupireddy mailto:bharath.rupireddyforpostg...@gmail.com>> wrote: On Tue, Jan 12, 2021 at 9:05 AM Amit Kapila

Re: Added schema level support for publication.

On Jan 11, 2021, at 5:06 PM, Bharath Rupireddy mailto:bharath.rupireddyforpostg...@gmail.com>> wrote: On Mon, Jan 11, 2021 at 1:29 PM japin mailto:japi...@hotmail.com>> wrote: Say a user has created a publication for a schema with hundreds of tables in it, at some point later, can he stop

Re: Terminate the idle sessions

-- Best regards Japin Li On Jan 7, 2021, at 10:03 AM, Thomas Munro mailto:thomas.mu...@gmail.com>> wrote: * I'm not entirely comfortable with the name "idle_session_timeout", because it sounds like it applies to all idle states, but actually it only applies when we're not in a transaction. I

Cannot ship records to subscriber for partition tables using logical replication (publish_via_partition_root=false)

Hi, hackers When I use logical stream replication on partition table, I find that if we create a new partitions after the subscription on subscriber, the records in new partitions cannot be shipped to the subscriber. Here is an example: 1. Create a view to check the subscription tables. ```

Re: Confused about stream replication protocol documentation

On Dec 23, 2020, at 8:11 PM, Fujii Masao mailto:masao.fu...@oss.nttdata.com>> wrote: On 2020/12/23 11:08, Li Japin wrote: On Dec 22, 2020, at 11:13 PM, Fujii Masao mailto:masao.fu...@oss.nttdata.com> <mailto:masao.fu...@oss.nttdata.com>> wrote: ‘B’ means a backend and ‘F

Re: Confused about stream replication protocol documentation

On Dec 22, 2020, at 11:13 PM, Fujii Masao mailto:masao.fu...@oss.nttdata.com>> wrote: ‘B’ means a backend and ‘F’ means a frontend. Maybe as [1] does, we should add the note like "Each is marked to indicate that it can be sent by a frontend (F) and a backend (B)" into the description about each

Confused about stream replication protocol documentation

Hi, all In Stream Replication Protocol [1], the documentation of `START_REPLICATION` message is XLogData (B) … Primary keepalive message (B) … Standby status update (F) … Hot Standby feedback message (F) ... I’m confused about the means of ‘B’ and ‘F’? If it doesn't make sense, why we

Re: Feature improvement for pg_stat_statements

Hi, On Dec 9, 2020, at 6:37 PM, Seino Yuki mailto:sein...@oss.nttdata.com>> wrote: 2020-12-01 01:04 に Fujii Masao さんは書きました: On 2020/11/30 23:05, Seino Yuki wrote: 2020-11-30 15:02 に Fujii Masao さんは書きました: On 2020/11/30 12:08, Seino Yuki wrote: 2020-11-27 22:39 に Fujii Masao さんは書きました: On

Re: About to add WAL write/fsync statistics to pg_stat_wal view

Hi, > On Dec 8, 2020, at 1:06 PM, Masahiro Ikeda wrote: > > Hi, > > I propose to add wal write/fsync statistics to pg_stat_wal view. > It's useful not only for developing/improving source code related to WAL > but also for users to detect workload changes, HW failure, and so on. > > I

Re: Printing LSN made easy

Hi, On Nov 30, 2020, at 9:06 PM, Ashutosh Bapat mailto:ashutosh.bapat@gmail.com>> wrote: On Fri, Nov 27, 2020 at 9:51 PM Li Japin mailto:japi...@hotmail.com>> wrote: Hi, Here, we cannot use sizeof(but) to get the buf size, because it is a pointer, so it always 8 bytes on

Re: Printing LSN made easy

Hi, Here, we cannot use sizeof(but) to get the buf size, because it is a pointer, so it always 8 bytes on 64-bit or 4 bytes on 32-bit machine. +char * +pg_lsn_out_buffer(XLogRecPtr lsn, char *buf) +{ + snprintf(buf, sizeof(buf), LSN_FORMAT, LSN_FORMAT_ARG(lsn)); + + return buf; +}

Re: Terminate the idle sessions

On Nov 24, 2020, at 11:20 PM, David G. Johnston mailto:david.g.johns...@gmail.com>> wrote: On Mon, Nov 23, 2020 at 11:22 PM Li Japin mailto:japi...@hotmail.com>> wrote: How about use “foreign-data wrapper” replace “postgres_fdw”? I don't see much value in avoiding mentioning t

Remove cache_plan argument comments to ri_PlanCheck

Hi, hackers I found that the cache_plan argument to ri_PlanCheck already been remove since 5b7ba75f7ff854003231e8099e3038c7e2eba875. I think we can remove the comments tor cache_plan to ri_PlanCheck. diff --git a/src/backend/utils/adt/ri_triggers.c b/src/backend/utils/adt/ri_triggers.c index

Re: Use macros for calculating LWLock offset

Thanks! On Nov 24, 2020, at 11:51 AM, Michael Paquier mailto:mich...@paquier.xyz>> wrote: On Fri, Nov 20, 2020 at 03:25:50PM +0900, Michael Paquier wrote: I agree that this makes this code a bit cleaner, so let's use those macros. Others may have some comments here, so let's wait a bit first.

Re: Terminate the idle sessions

Hi, David Thanks for your suggestion! On Nov 24, 2020, at 11:39 AM, David G. Johnston mailto:david.g.johns...@gmail.com>> wrote: On Mon, Nov 23, 2020 at 5:02 PM kuroda.hay...@fujitsu.com mailto:kuroda.hay...@fujitsu.com>> wrote: No one have any comments,

Re: Terminate the idle sessions

Hi, Kuroda Thank for your review. > On Nov 24, 2020, at 8:01 AM, kuroda.hay...@fujitsu.com wrote: > > No one have any comments, patch tester says OK, and I think this works well. > I changed status to "Ready for Committer." > > Hayato Kuroda > FUJITSU LIMITED > > -Original Message- >

Re: Terminate the idle sessions

On Nov 18, 2020, at 2:22 PM, kuroda.hay...@fujitsu.com wrote: Oops.. I forgot putting my suggestion. Sorry. How about substituting sigalrm_delivered to true in the reschedule_timeouts()? Maybe this processing looks strange, so some comments should be put too.

Re: Terminate the idle sessions

On Nov 18, 2020, at 10:40 AM, kuroda.hay...@fujitsu.com wrote: I’m not familiar with the system interrupt, however, the sigalrm_due_at is subsutitue between HOLD_INTERRUPTS() and RESUM_INTERRUPTS(), so I think it cannot be interrupted. The following comments

Re: Terminate the idle sessions

On Nov 17, 2020, at 2:07 PM, kuroda.hay...@fujitsu.com wrote: Dear Li, David, > Additionally, using postgres_fdw within the server doesn't cause issues, > its using postgres_fdw and the remote server having this setting set to zero > that causes a problem. I

Re: Terminate the idle sessions

On Nov 17, 2020, at 10:53 AM, David G. Johnston mailto:david.g.johns...@gmail.com>> wrote: On Monday, November 16, 2020, Li Japin mailto:japi...@hotmail.com>> wrote: Consider setting this for specific users instead of as a server default. Client connections managed by connec

Re: Terminate the idle sessions

-- Best regards Japin Li On Nov 17, 2020, at 7:59 AM, David G. Johnston mailto:david.g.johns...@gmail.com>> wrote: On Mon, Nov 16, 2020 at 5:41 AM Li Japin mailto:japi...@hotmail.com>> wrote: Thanks for your review! Attached. Reading the doc changes: I'd rather not name

Re: Terminate the idle sessions

Hi Kuroda, On Nov 16, 2020, at 1:22 PM, kuroda.hay...@fujitsu.com wrote: @@ -30,6 +30,7 @@ typedef enum TimeoutId STANDBY_DEADLOCK_TIMEOUT, STANDBY_TIMEOUT, STANDBY_LOCK_TIMEOUT, + IDLE_SESSION_TIMEOUT, IDLE_IN_TRANSACTION_SESSION_TIMEOUT, /* First

Re: Terminate the idle sessions

On Nov 13, 2020, at 6:27 PM, kuroda.hay...@fujitsu.com wrote: I read your patch, and I think the documentation is too simple to avoid all problems. (I think if some connection pooling is used, the same problem will occur.) Could you add some explanations in

Re: Tab complete for alter table rls

Thanks Michael! -- Best regards Japin Li > On Oct 24, 2020, at 9:49 AM, Michael Paquier wrote: > > On Fri, Oct 23, 2020 at 04:37:18PM +0900, Michael Paquier wrote: >> No worries. Good catch. I'll try to test that and apply it later, >> but by reading the code it looks like you got that

Tab complete for alter table rls

Sorry, I forgot add the subject. -- Best regards Japin Li On Oct 23, 2020, at 1:19 PM, Li Japin mailto:japi...@hotmail.com>> wrote: Hi, hackers I find that ALTER TABLE xxx FORCE/NO FORCE ROW LEVEL SECURITY cannot support tab complete. The attached add the tab complete for rls. diff

[no subject]

Hi, hackers I find that ALTER TABLE xxx FORCE/NO FORCE ROW LEVEL SECURITY cannot support tab complete. The attached add the tab complete for rls. diff --git a/src/bin/psql/tab-complete.c b/src/bin/psql/tab-complete.c index 561fe1dff9..b2b4f1fd4d 100644 --- a/src/bin/psql/tab-complete.c +++

Re: Enumize logical replication message actions

> On Oct 16, 2020, at 3:25 PM, Ashutosh Bapat > wrote: > > Hi All, > Logical replication protocol uses single byte character to identify > different chunks of logical repliation messages. The code uses > character literals for the same. These literals are used as bare > constants in code as

Re: Remove unnecessary else branch

On Oct 13, 2020, at 9:59 PM, Hamid Akhtar mailto:hamid.akh...@gmail.com>> wrote: On Tue, Oct 13, 2020 at 6:37 PM Heikki Linnakangas mailto:hlinn...@iki.fi>> wrote: On 13/10/2020 16:30, Li Japin wrote: > Hi, > > I found in guc-file.l we can omit the else branch in A

Re: Remove unnecessary else branch

On Oct 13, 2020, at 9:59 PM, Hamid Akhtar mailto:hamid.akh...@gmail.com>> wrote: On Tue, Oct 13, 2020 at 6:37 PM Heikki Linnakangas mailto:hlinn...@iki.fi>> wrote: On 13/10/2020 16:30, Li Japin wrote: > Hi, > > I found in guc-file.l we can omit the else branch in A

Remove unnecessary else branch

Hi, I found in guc-file.l we can omit the else branch in AbsoluteConfigLocation(). diff --git a/src/backend/utils/misc/guc-file.l b/src/backend/utils/misc/guc-file.l index c98e220295..9d4b3d7236 100644 --- a/src/backend/utils/misc/guc-file.l +++ b/src/backend/utils/misc/guc-file.l @@ -522,23

Re: Optimize memory allocation code

On Sep 29, 2020, at 9:30 PM, Alvaro Herrera mailto:alvhe...@2ndquadrant.com>> wrote: On 2020-Sep-26, Li Japin wrote: Thanks! How big is this overhead? Is there any way I can test it？ You could also have a look at the assembly code that your compiler generates -- particularly exami

Re: Optimize memory allocation code

> On Sep 26, 2020, at 8:09 AM, Julien Rouhaud wrote: > > Hi, > > On Sat, Sep 26, 2020 at 12:14 AM Li Japin wrote: >> >> Hi, hackers! >> >> I find the palloc0() is similar to the palloc(), we can use palloc() inside >> palloc0() >&g

Optimize memory allocation code

Hi, hackers! I find the palloc0() is similar to the palloc(), we can use palloc() inside palloc0() to allocate space, thereby I think we can reduce duplication of code. Best regards! -- Japin Li 0001-Optimize-memory-allocation-code.patch Description:

Re: Parallelize stream replication process

Thanks for your advice! This help me a lot. > On Sep 17, 2020, at 9:18 PM, Jakub Wartak wrote: > > Li Japin wrote: > >> If we can improve the efficiency of replay, then we can shorten the database >> recovery time (streaming replication or database crash recovery).

Re: Parallelize stream replication process

> On Sep 15, 2020, at 3:41 PM, Fujii Masao wrote: > > > > On 2020/09/15 13:41, Bharath Rupireddy wrote: >> On Tue, Sep 15, 2020 at 9:27 AM Li Japin wrote: >>> >>> For now, postgres use single process to send, receive and replay the WAL >>>

Re: Parallelize stream replication process

Thanks for clarifying the questions! > On Sep 15, 2020, at 12:41 PM, Bharath Rupireddy > wrote: > > On Tue, Sep 15, 2020 at 9:27 AM Li Japin wrote: >> >> For now, postgres use single process to send, receive and replay the WAL >> when we use stream repl

Parallelize stream replication process

Hi, hackers For now, postgres use single process to send, receive and replay the WAL when we use stream replication, is there any point to parallelize this process? If it does, how do we start? Any thoughts? Best regards Japin Li

Re: Docs: inaccurate description about config settings

On Sep 1, 2020, at 8:20 PM, Ian Lawrence Barwick mailto:barw...@gmail.com>> wrote: 2020年9月1日(火) 19:37 Li Japin mailto:japi...@hotmail.com>>: Hi, hackers When I setup a stream replication I found that the documentation says that promote_trigger_file parameter can

Docs: inaccurate description about config settings

Hi, hackers When I setup a stream replication I found that the documentation says that promote_trigger_file parameter can only be set in the postgresql.conf file or on the server command line, however, it can also be put into postgresql.auto.conf. If I use include to import a new config, it

Re: Terminate the idle sessions

> On Aug 31, 2020, at 11:43 AM, Thomas Munro wrote: > > On Mon, Aug 31, 2020 at 2:40 PM Li Japin wrote: >> Could you give the more details about the test instructions? > > Hi Japin, > > Sure. Because I wasn't trying to get reliable TPS number or anything, > I

Re: Terminate the idle sessions

On Aug 31, 2020, at 8:51 AM, Thomas Munro mailto:thomas.mu...@gmail.com>> wrote: The main problem I have with it is the high frequency setitimer() calls. If you enable both statement_timeout and idle_session_timeout, then we get up to huge number of system calls, like the following strace -c

Re: file_fdw vs relative paths

On Aug 25, 2020, at 8:26 AM, Bruce Momjian mailto:br...@momjian.us>> wrote: Yes, I tested back to 9.5 too: CREATE EXTENSION file_fdw; CREATE SERVER pgconf FOREIGN DATA WRAPPER file_fdw; CREATE FOREIGN TABLE pgconf (line TEXT) SERVER pgconf OPTIONS ( filename 'postgresql.conf', format 'text',

Re: Terminate the idle sessions

On Aug 18, 2020, at 9:19 AM, Kyotaro Horiguchi mailto:horikyota@gmail.com>> wrote: The same already happens for idle_in_transaction_session_timeout and we can use "ALTER ROLE/DATABASE SET" to dislable or loosen them, it's a bit cumbersome, though. I don't think we should (at least

Re: Terminate the idle sessions

On Aug 14, 2020, at 2:15 PM, Bharath Rupireddy mailto:bharath.rupireddyforpostg...@gmail.com>> wrote: I think, since the idle_session_timeout is by default disabled, we have no problem. My thought is what if a user enables the feature(knowingly or unknowingly) on the remote backend? If the

Re: Terminate the idle sessions

Hi, On Aug 11, 2020, at 5:42 AM, Cary Huang mailto:cary.hu...@highgo.ca>> wrote: I applied this patch to the PG13 branch and generally this feature works as described. The new "idle_session_timeout" that controls the idle session disconnection is not in the default postgresql.conf and I think

Re: Terminate the idle sessions

On Jun 10, 2020, at 10:27 PM, Adam Brusselback mailto:adambrusselb...@gmail.com>> wrote: My use case is, I have a primary application that connects to the DB, most users work through that (setting is useless for this scenario, app manages it's connections well enough). I also have a number

Re: Terminate the idle sessions

On Jun 10, 2020, at 4:25 PM, Michael Paquier mailto:mich...@paquier.xyz>> wrote: Idle sessions staying around can be a problem in the long run as they impact snapshot building. You could for example use a background worker to do this work, like that:

Re: Terminate the idle sessions

On Jun 9, 2020, at 10:35 PM, David G. Johnston mailto:david.g.johns...@gmail.com>> wrote: I’m curious as to the use case because I cannot imagine using this. Idle connections are normal. Seems better to monitor them and conditionally execute the disconnect backend function from the

Terminate the idle sessions

Hi, hackers When some clients connect to database in idle state, postgres do not close the idle sessions, here i add a new GUC idle_session_timeout to let postgres close the idle sessions, it samilar to idle_in_transaction_session_timeout. Best, regards. Japin Li

Code cleanup for build_regexp_split_result

Hi hackers, I find the build_regexp_split_result() has redundant codes, we can move it to before the condition check, can we? Best regards. Japin Li 0001-Code-cleanup-for-build_regexp_split_result.patch Description: 0001-Code-cleanup-for-build_regexp_split_result.patch

Re: Duplicate function call on timestamp2tm

Thanks for your confirm. Is there anything I can do? On Dec 12, 2019, at 11:13 PM, Tom Lane mailto:t...@sss.pgh.pa.us>> wrote: Ah, after looking in the git history, not quite that ancient: this duplication dates to commit 258ee1b63, which moved these switch cases from the "if (type == RESERV)"

Duplicate function call on timestamp2tm

Hi, I find there is a duplicate function call on timestamp2tm in timestamptz_part and timestamp_part. Is that necessary? I remove the latter one and it also works. Best, Japin. remove-duplicate-timestamp2tm-function-call.patch Description: remove-duplicate-timestamp2tm-function-call.patch

Re: Built-in connection pooler

Hi, Konstantin I test the patch-16 on postgresql master branch, and I find the temporary table cannot removed when we re-connect to it. Here is my test: japin@ww-it:~/WwIT/postgresql/Debug/connpool$ initdb The files belonging to this database system will be owned by user "japin". This user must

Re: Symbol referencing errors

, Li Japin wrote: Yes, those errors does not impact the postgresql, but when I use oracle_fdw extension, I couldn't startup the postgresql, and I find that the dlopen throw an error which lead postmaster exit, and there is not more information. That may wall be a bug in oracle_fdw, since I have

Re: Symbol referencing errors

On 4/23/19 12:09 PM, Tom Lane wrote: > AFAICT they're harmless, so my advice is just ignore them. > > If you're sufficiently annoyed by them to find the cause > and try to fix it, go ahead, but I haven't heard anyone > else worried about it. It might be that SmartOS wants > something like what

Symbol referencing errors

Hi, When I compile PostgreSQL-11.2 on SmartOS, I find the following errors: Undefined            first referenced  symbol                  in file per_MultiFuncCall   adminpack.o end_MultiFuncCall   adminpack.o BuildTupleFromCStrings  adminpack.o