{"payload":{"feedbackUrl":"https://github.com/orgs/community/discussions/53140","repo":{"id":646423,"defaultBranch":"master","name":"pgbouncer","ownerLogin":"pgbouncer","currentUserCanPush":false,"isFork":false,"isEmpty":false,"createdAt":"2010-05-04T10:39:40.000Z","ownerAvatar":"https://avatars.githubusercontent.com/u/11858179?v=4","public":true,"private":false,"isOrgOwned":true},"refInfo":{"name":"","listCacheKey":"v0:1709550670.0","currentOid":""},"activityList":{"items":[{"before":"817c20a97e27bb3958a348f5bd9d39688a60e7c0","after":"d8eee4be3cf43106dedf1e08d15cc4ca6a479efd","ref":"refs/heads/master","pushedAt":"2024-06-05T08:42:16.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Change systemd service file to automatically restart PgBouncer (#1080)\n\nIn production environments most people will want PgBouncer to\r\nautomatically restart if it crashes for some reason. This changes our\r\nsystemd template file to do that.\r\n\r\nAs suggested by @phemmer in\r\nhttps://github.com/pgbouncer/pgbouncer/issues/314#issuecomment-2129664358","shortMessageHtmlLink":"Change systemd service file to automatically restart PgBouncer (#1080)"}},{"before":"6727c6ad7e722458ed9ec1e054093eedffc3c86d","after":"817c20a97e27bb3958a348f5bd9d39688a60e7c0","ref":"refs/heads/master","pushedAt":"2024-05-28T22:09:07.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Improve documentation on connecting to AWS RDS and SCRAM (#1083)\n\nSome managed servers such as AWS RDS for PostgreSQL restrict access to\r\n`pg_authid` (`pg_shadow`) which makes configuration of authentication\r\nbetween PgBouncer and backend server somewhat difficult. Configuring\r\nSCRAM for both PgBouncer and backend server has been clearly a pain\r\npoint for some PgBouncer users who experienced connection problems\r\nrelated to failing authentication (see #787, #981).\r\n\r\nThis includes a description of this problem in the docs and explains\r\nhow a user can work around it. \r\n\r\nResolves #787\r\nResolves #981","shortMessageHtmlLink":"Improve documentation on connecting to AWS RDS and SCRAM (#1083)"}},{"before":"f06a8cbb70efe31371a483ffc5f50b7b1d4526c6","after":"6727c6ad7e722458ed9ec1e054093eedffc3c86d","ref":"refs/heads/master","pushedAt":"2024-05-27T06:49:57.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Fix minor details in config.md (#1082)","shortMessageHtmlLink":"Fix minor details in config.md (#1082)"}},{"before":"788517e2032f11ff851102f3606fef3b977da401","after":"f06a8cbb70efe31371a483ffc5f50b7b1d4526c6","ref":"refs/heads/master","pushedAt":"2024-05-25T07:02:14.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Add support for \"all\" address type in HBA file (#1078)\n\nThis change adds support for \"all\" address type in HBA file that matches\r\nboth 0.0.0.0/0 and ::/0, similar to PostgreSQL HBA.\r\n\r\nhttps://postgresql.org/docs/16/auth-pg-hba-conf.html#:~:text=You%20can%20also%20write%20all%20to%20match%20any%20IP%20address","shortMessageHtmlLink":"Add support for \"all\" address type in HBA file (#1078)"}},{"before":"34ab8180c9a0a8791bf1e122b07b857c8409f3aa","after":"788517e2032f11ff851102f3606fef3b977da401","ref":"refs/heads/master","pushedAt":"2024-05-23T06:55:03.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"petere","name":"Peter Eisentraut","path":"/petere","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/105543?s=80&v=4"},"commit":{"message":"Increase c-ares version requirement to 1.9.0\n\nThis allows us to drop our own implementation of ares_parse_soa_reply().\n\nSee c-ares/c-ares@979bf951d and c-ares/c-ares@cc067c2.\n\nThis version is now available in all relevant operating system\nreleases.","shortMessageHtmlLink":"Increase c-ares version requirement to 1.9.0"}},{"before":"6dd2aaf764a74289871a051c9d5eaeae9da0d19d","after":"34ab8180c9a0a8791bf1e122b07b857c8409f3aa","ref":"refs/heads/master","pushedAt":"2024-05-20T06:41:48.000Z","pushType":"pr_merge","commitsCount":2,"pusher":{"login":"petere","name":"Peter Eisentraut","path":"/petere","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/105543?s=80&v=4"},"commit":{"message":"ci: Use venv on Alpine\n\nApply the changes from 8ec53d2 to the Alpine task as well. It is now\nrequired there.","shortMessageHtmlLink":"ci: Use venv on Alpine"}},{"before":"8d16ca3cbac14132997dd411e05e826eb32e4bf4","after":"6dd2aaf764a74289871a051c9d5eaeae9da0d19d","ref":"refs/heads/master","pushedAt":"2024-05-16T16:02:26.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"The 1st argument of calloc should be 'nelem'. (#1073)\n\nThe 1st argument of calloc is the number of elements of the array that\r\nwe want to allocate space for[^1]. This patch is trying to pgbouncer's\r\nCI happy.\r\n\r\n[^1]: https://man7.org/linux/man-pages/man3/calloc.3p.html","shortMessageHtmlLink":"The 1st argument of calloc should be 'nelem'. (#1073)"}},{"before":"59f35332a2edbff74ad18d4a5be93723fa078de2","after":"8d16ca3cbac14132997dd411e05e826eb32e4bf4","ref":"refs/heads/master","pushedAt":"2024-05-16T14:19:39.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Make avg_wait_time a meaningful metric (#727)\n\nCurrently avg_wait_time is calculated as the wait time accumulated by\r\nclients that were assigned a backend in the current stats_period and\r\naveraged over the time in that period. That means that a connection\r\nthat's been waiting for 10 minutes will cause (assuming the default\r\nstats_period of 60s) a reported 10s average wait time in the\r\nstats_period it is assigned a connection (10m = 600s then 600s / 60s).\r\nIt's fairly obvious that's not meaningful, because the reported value\r\nhas no relation to the wait time experienced by an average client.\r\n\r\nIf we instead keep track of how many times backends have been assigned\r\nfor the current pool (if we were limited to transaction pooling, we\r\ncould use transaction count, for example, but we want something that's\r\ngeneric across all pooling modes) then we can report the wait time\r\nexperienced on average by a client assigned a backend during the current\r\nstats period.\r\n\r\nThis also removes some dead code (the writing to `st_total` and `old_total`).\r\n\r\nFixes #316\r\n\r\n---------\r\n\r\nCo-authored-by: jcoleman \r\nCo-authored-by: Jelte Fennema-Nio ","shortMessageHtmlLink":"Make avg_wait_time a meaningful metric (#727)"}},{"before":"de33cd058a6297f063e0ba27612c093fa989de30","after":"59f35332a2edbff74ad18d4a5be93723fa078de2","ref":"refs/heads/master","pushedAt":"2024-05-10T17:44:57.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Support --config=value format in options startup parameter (#1064)\n\nThe `-c config=value` pattern has been supported in PgBouncer since\r\nPR #878. As described in #952 Postgres also allows to use the\r\n`--config=value`, which honestly seems more readable, but for some\r\nreason is mentioned almost nowhere in the Postgres documentation.\r\n\r\nSince #878 this always throws an error, and putting `options` in\r\n`ignore_startup_parameters` wouldn't get rid of this error. So this\r\nshould actually be considered a regression for the same reason as #907\r\nwas considered a regression (although a much less impactful one given\r\nwe've only received two reports of this causing issues).\r\n\r\nFixes #952","shortMessageHtmlLink":"Support --config=value format in options startup parameter (#1064)"}},{"before":"e1c6c2f2269539ae05c3f2a6fb03506340d90102","after":"de33cd058a6297f063e0ba27612c093fa989de30","ref":"refs/heads/master","pushedAt":"2024-05-06T14:38:14.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Redesign shutdown logic to support rolling restarts using so_reuseport (#902)\n\nIn 1.20 we announced deprecation of online restarts using the `-R` flag\r\nand recommended for people to use `so_reuseport` instead. But the safe\r\nshutdown logic that we executed when receiving a `SIGINT` signal didn't\r\nactually allow for zero-downtime rolling restarts.\r\n\r\nThis PR changes the behaviour of our shutdown signals to allow for\r\nrolling restarts. The changes that this PR makes are:\r\n\r\n1. SIGTERM does not do \"immediate shutdown\" anymore, but instead\r\ntriggers a newly introduced \"super safe shutdown\". This \"super safe\r\nshutdown\" waits for all clients to disconnect before shutting down the\r\nserver. This can be used for rolling restarts as described in the newly\r\nadded documentation. A second SIGTERM will cause an \"immediate\r\nshutdown\".\r\n2. SIGINT mostly keeps doing what it was doing previously: waiting until\r\nall servers are released before shutting down PgBouncer. But it\r\nincorporates a few improvements from the new SIGTERM logic:\r\n1. It also stops listening for new connections while PgBouncer is\r\nwaiting for servers to be released. This does have the downside that\r\nSIGUSR2 cannot be used to cancel a safe shutdown anymore, but that was\r\nquite a weird/niche feature anyway.\r\n2. Clients that have released their server and then do another query\r\nwill be disconnected, instead of putting them in pause mode until either\r\n`query_wait_timeout` triggers or PgBouncer shuts down. This way they can\r\nreconnect quicker to another PgBouncer, because they were never going to\r\nget a new server anyway.\r\n 3. A that a second SIGINT will trigger an \"immediate shutdown\"\r\n3. SIGQUIT handling is introduced and now causes an immediate shutdown.\r\n\r\nSince this changes the existing behaviour of SIGTERM and SIGINT, this\r\ncan be considered a breaking change. But the amount of breakage should\r\nbe minimal since most of the time people would not want an immediate\r\nshutdown, and the new SIGINT behaviour is closer to the behaviour most\r\npeople would expect.\r\n\r\nRelated to #894\r\nRelated to #900\r\n\r\nSince this changes SIGTERM to not do a fast shutdown anymore this also\r\nindirectly addresses #806 and #770.\r\n\r\nFixes #806 \r\nFixes #770","shortMessageHtmlLink":"Redesign shutdown logic to support rolling restarts using so_reuseport ("}},{"before":"3f6b3dcece7bad9f418be1d3060a47adc854e389","after":"e1c6c2f2269539ae05c3f2a6fb03506340d90102","ref":"refs/heads/master","pushedAt":"2024-05-06T14:09:10.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Support replication connections through PgBouncer (#876)\n\nIn session pooling mode PgBouncer is pretty much a transparent proxy,\r\ni.e. the client does normally not even need to know that PgBouncer is in\r\nthe middle. This allows things like load balancing and failovers without\r\nthe client needing to know about this at all. But as soon as replication\r\nconnections are needed, this was not possible anymore, because PgBouncer\r\nwould reject those instead of proxying them to the right server.\r\n\r\nThis PR fixes that by also proxying replication connections. They are\r\nhandled pretty differently from normal connections though. A client and\r\nserver replication connection will form a strong pair, as soon as one is\r\nclosed the other is closed too. So, there's no caching of the server\r\nreplication connections, like is done for regular connections. Reusing\r\nreplication connections comes with a ton of gotchas. Postgres will\r\nthrow errors in many cases when trying to do so. So simply not\r\ndoing it seems like a good tradeoff for ease of implementation.\r\nEspecially because replication connections are pretty much always\r\nvery long lived. So re-using them gains pretty much no performance\r\nbenefits.\r\n\r\nFixes #382\r\n\r\nDepends on #945","shortMessageHtmlLink":"Support replication connections through PgBouncer (#876)"}},{"before":"f7ffd67ed8065ade25f375abeecad4a73eb373b4","after":"3f6b3dcece7bad9f418be1d3060a47adc854e389","ref":"refs/heads/master","pushedAt":"2024-05-06T09:57:28.000Z","pushType":"pr_merge","commitsCount":4,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Add support for negotiating the protocol version with the client (#1007)\n\nPgBouncer currently only supports protocol version 3.0. However, it was\r\nstill accepting connections if the client asked for the 3.1 protocol\r\nversion. This is obviously a bug, because PgBouncer likely won't\r\nsupport the changes introduced in 3.1 of the protocol without\r\nany code changes.\r\n\r\nThis starts sending the NegotiateProtocolVersion message when the client\r\nasks for a minor version that PgBouncer does not support. Also, instead\r\nof erroring when it receives unsupported protocol extensions parameters\r\nit will include these in this NegotiateProtocolVersion message.\r\n\r\nFor reference this is the Postgres commit introducing\r\nNegotiateProtocolVersion:\r\nhttps://github.com/postgres/postgres/commit/ae65f6066dc3d19a55f4fdcd3b30003c5ad8dbed\r\n\r\nI tested this manually using this patchset that bumps the protocol\r\nversion.\r\nhttps://www.postgresql.org/message-id/flat/CAGECzQScQ3N-Ykv2j4NDyDtrPPc3FpRoa%3DLZ-2Uj2ocA4zr%3D4Q%40mail.gmail.com#cd9e8407820d492e8f677ee6a67c21ce","shortMessageHtmlLink":"Add support for negotiating the protocol version with the client (#1007)"}},{"before":"6c629ab98e8c06aec05feda49019908693121755","after":"f7ffd67ed8065ade25f375abeecad4a73eb373b4","ref":"refs/heads/master","pushedAt":"2024-05-06T08:04:23.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Update libusual (#1063)\n\nFixes #713\r\nFixes #842","shortMessageHtmlLink":"Update libusual (#1063)"}},{"before":"e0d5694543b5b508c55b07bb8d71361a429e03e3","after":"6c629ab98e8c06aec05feda49019908693121755","ref":"refs/heads/master","pushedAt":"2024-05-03T00:07:51.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"eulerto","name":"Euler Taveira","path":"/eulerto","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/213604?s=80&v=4"},"commit":{"message":"ci: install PAM development package (#1061)\n\nAlthough --with-pam option has been used, since the PAM development\r\nfiles are not installed, it was not building with PAM support.\r\n\r\nIt was the reason that the CI does not detect an issue with the commit\r\nb003f7b8b9dbea6a2b95980dc66bd1bd7da964aa.","shortMessageHtmlLink":"ci: install PAM development package (#1061)"}},{"before":"4da2a672bae236bdc12405bdb0e5a9f7733a5beb","after":"e0d5694543b5b508c55b07bb8d71361a429e03e3","ref":"refs/heads/master","pushedAt":"2024-05-02T23:34:43.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"eulerto","name":"Euler Taveira","path":"/eulerto","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/213604?s=80&v=4"},"commit":{"message":"PAM: fix user management refactor (#1060)\n\nA previous commit (b003f7b8b9dbea6a2b95980dc66bd1bd7da964aa) that\r\nrefactored the user management forgot to change PAM code.","shortMessageHtmlLink":"PAM: fix user management refactor (#1060)"}},{"before":"b003f7b8b9dbea6a2b95980dc66bd1bd7da964aa","after":"4da2a672bae236bdc12405bdb0e5a9f7733a5beb","ref":"refs/heads/master","pushedAt":"2024-05-02T08:56:11.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Correctly handle large and partial startup packets (#1058)\n\nWe were bailing out from a connection if we did not receive the full\r\nstartup packet in one go, or if it didn't fit in pkt_buf. This fixes\r\nthese issues by reusing the packet buffering logic that was introduced\r\nfor prepared statement support for our startup packet parsing.\r\n\r\nFixes #1056","shortMessageHtmlLink":"Correctly handle large and partial startup packets (#1058)"}},{"before":"25e5f034f746faceff13ddb95ac30d18daa71bc7","after":"b003f7b8b9dbea6a2b95980dc66bd1bd7da964aa","ref":"refs/heads/master","pushedAt":"2024-05-02T08:43:16.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Refactor user management (#1052)\n\nBased upon #1039, but changed a lot by splitting the PgUser type into\r\ntwo different types. One type for storing global user configuration, auth\r\ninfo, and connection count tracking and a second type for dynamically created\r\nusers (which always references a global user).\r\n\r\nThis makes user configuration work for dynamically created users and\r\nwill also\r\nlist dynamically created users in the output of `SHOW USERS`.\r\n\r\nFixes #484\r\nFixes #706\r\nResolves #1039\r\n\r\nCo-authored-by: Ben Chobot \r\nCo-authored-by: benchub ","shortMessageHtmlLink":"Refactor user management (#1052)"}},{"before":"e18f60074c0558646b5db95ff2461d1e03a3dfc5","after":"25e5f034f746faceff13ddb95ac30d18daa71bc7","ref":"refs/heads/master","pushedAt":"2024-04-30T22:07:08.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Add `server_lifetime` to databases config (#1057)\n\nThis PR makes it possible to specify `server_lifetime` on a per database\r\nbasis in the `[databases]` section.\r\n\r\nThis can be useful when a pgbouncer instance that pools connections\r\nfor several different severs, of which some makes heavy use of partitioning.\r\nThe partitioning causes connections to become very memory heavy\r\novertime and by lowering server_lifetime these connections can be refreshed\r\nmore often. Without this change that could only be done by lowering\r\nserver_lifetime globally and thus incurring the overhead of re-authenticating\r\neven for servers where no partitioning is used.","shortMessageHtmlLink":"Add server_lifetime to databases config (#1057)"}},{"before":"322acfbb7d8342f9ebcfa2734cb9129b8a938d2a","after":"e18f60074c0558646b5db95ff2461d1e03a3dfc5","ref":"refs/heads/master","pushedAt":"2024-04-25T09:47:57.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Allow pool_size per user (#1049)\n\nSpecifying `pool_size` is now possible in section `[users]` beside\r\n`[databases]` or as a global param. Possibility of specifying\r\n`pool_size` in user configuration was requested in the associated issue.\r\nResolves #911.","shortMessageHtmlLink":"Allow pool_size per user (#1049)"}},{"before":"8544b49574848a93252749c72a693aba576e5963","after":"322acfbb7d8342f9ebcfa2734cb9129b8a938d2a","ref":"refs/heads/master","pushedAt":"2024-04-17T16:14:37.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Improves README (#1051)\n\nThe README uses ... in the configure command line to indicate that\r\noptions can be informed. However, it can confuse people (like in the\r\nreferred issue #1050). Removes the 3 dots and explains that configure\r\noptions can be informed.\r\n\r\nSince I'm at it, informs that uthash submodule is also fetched if you\r\nare following the Git instructions.\r\n\r\nFixes #1050","shortMessageHtmlLink":"Improves README (#1051)"}},{"before":"275c669110b8190609bce19890bc4186ff92ac2f","after":"8544b49574848a93252749c72a693aba576e5963","ref":"refs/heads/master","pushedAt":"2024-04-10T12:10:23.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Fix tests on non US datestyle configured systems (#1047)\n\nThis fixes an issue that I ran into where\r\n`test_equivalent_startup_param` failed on my new laptop because initdb\r\nwould use `'iso, dmy'` as datestyle instead of `'iso, mdy'`.","shortMessageHtmlLink":"Fix tests on non US datestyle configured systems (#1047)"}},{"before":"0cf83fa909e37cfa8e18370a31ebbb7f9ff3243c","after":"275c669110b8190609bce19890bc4186ff92ac2f","ref":"refs/heads/master","pushedAt":"2024-04-10T08:50:12.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Silence deprecation warnings of c-ares (#1046)\n\nThe newest version of c-ares has started adding deprecation warnings.\r\nLet's silence those for now to make MacOS builds pass again.\r\n\r\nWe might want to actually start using the new APIs in the future, but\r\nsince we want to support some fairly old systems I'm not sure that's a\r\ngood plan now. And given [the library author stated that the old\r\nfunctions would probably never be removed][1], I think we're fine with\r\nsimply ignoring the warnings.\r\n\r\n[1]: https://github.com/c-ares/c-ares/pull/732#issuecomment-2028454381","shortMessageHtmlLink":"Silence deprecation warnings of c-ares (#1046)"}},{"before":"b40112755c2a637c896c0bd067edfb5e5624d53c","after":"0cf83fa909e37cfa8e18370a31ebbb7f9ff3243c","ref":"refs/heads/master","pushedAt":"2024-04-10T08:41:10.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Add outstanding request for auth_query (#1034)\n\nI realized I was a bit too trigger-happy when removing all references to\r\n`expected_rfq_count` in #1025. One occurrence should have been replaced\r\nwith adding an outstanding request. Normally this isn't an issue, but it\r\ncould cause the server connection not to be closed if the client\r\ndisconnects before the auth_query is finished running.\r\n\r\nTo be clear, #1025 did not introduce a regression in this respect.\r\nBecause `expect_rfq_count` being non-zero on would not cause\r\na server connection closure when the server was released.","shortMessageHtmlLink":"Add outstanding request for auth_query (#1034)"}},{"before":"b1ceb1ccb3378152e055a1b5471fb1cc03e4dc58","after":"b40112755c2a637c896c0bd067edfb5e5624d53c","ref":"refs/heads/master","pushedAt":"2024-04-10T08:39:23.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Fix typos (#1045)\n\nFound via `codespell -L ser,ue,parms` and `typos --hidden --format\r\nbrief`","shortMessageHtmlLink":"Fix typos (#1045)"}},{"before":"0401b9f1b82d446f67e3d21084ef70ce0ab57f93","after":"b1ceb1ccb3378152e055a1b5471fb1cc03e4dc58","ref":"refs/heads/master","pushedAt":"2024-03-14T11:17:02.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Change SHOW USERS to list connections (#1040)\n\nWhile reviewing #1039 I had a hard time seeing if\r\nmax_user_connections was applied correctly because `SHOW USERS` didn't\r\nlist it. This addresses that by adding a `max_user_connections` and\r\n`current_connections` column to the output of `SHOW USERS`.","shortMessageHtmlLink":"Change SHOW USERS to list connections (#1040)"}},{"before":"eec906c30ddabc5676aa6ef42704db787314d617","after":"0401b9f1b82d446f67e3d21084ef70ce0ab57f93","ref":"refs/heads/master","pushedAt":"2024-03-13T18:37:18.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"Expand code comment on waiting_cancel_req_list (#1016)\n\nThis clarifies why we have waiting_cancel_req_list in addition to\r\nwaiting_client_list.","shortMessageHtmlLink":"Expand code comment on waiting_cancel_req_list (#1016)"}},{"before":"5661fc2119a32beaec1f14dd1107fcc1430d79c7","after":"eec906c30ddabc5676aa6ef42704db787314d617","ref":"refs/heads/master","pushedAt":"2024-03-11T08:35:48.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"docs: Mention that pem certs / keys are usable by pgbouncer (#1036)\n\nFixes #1035","shortMessageHtmlLink":"docs: Mention that pem certs / keys are usable by pgbouncer (#1036)"}},{"before":null,"after":"f81e4ee952cfacffceae355b15c0016bf7e4f25e","ref":"refs/heads/revert-996-hba-pgident-map","pushedAt":"2024-03-04T11:11:10.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"emelsimsek","name":"Emel Şimşek","path":"/emelsimsek","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/13130350?s=80&v=4"},"commit":{"message":"Revert \"Introduce auth_ident_file setting and enhance cert and peer auth methods with user name maps\"","shortMessageHtmlLink":"Revert \"Introduce auth_ident_file setting and enhance cert and peer a…"}},{"before":"de1e9cb75324a624d18fc3dd5d714d12a8da0db6","after":"5661fc2119a32beaec1f14dd1107fcc1430d79c7","ref":"refs/heads/master","pushedAt":"2024-03-04T10:32:51.000Z","pushType":"pr_merge","commitsCount":30,"pusher":{"login":"emelsimsek","name":"Emel Şimşek","path":"/emelsimsek","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/13130350?s=80&v=4"},"commit":{"message":"Introduce auth_ident_file setting and enhance cert and peer auth methods with user name maps (#996)\n\nPostgres supports user name maps[1] for `cert` and `peer` authentication\r\nmethods. This change provides the same functionality with some\r\nlimitations for pgbouncer. The limitations are listed in \"Ident map file\r\nformat\" section of the docs.\r\n\r\n1. https://www.postgresql.org/docs/current/auth-username-maps.html\r\n\r\nFixes #767\r\nFixes #148","shortMessageHtmlLink":"Introduce auth_ident_file setting and enhance cert and peer auth meth…"}},{"before":"ba96807044b403e5612e57265eca991e3827c925","after":"de1e9cb75324a624d18fc3dd5d714d12a8da0db6","ref":"refs/heads/master","pushedAt":"2024-03-04T03:11:20.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"JelteF","name":"Jelte Fennema-Nio","path":"/JelteF","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1162278?s=80&v=4"},"commit":{"message":"v1.22.1","shortMessageHtmlLink":"v1.22.1"}}],"hasNextPage":true,"hasPreviousPage":false,"activityType":"all","actor":null,"timePeriod":"all","sort":"DESC","perPage":30,"cursor":"djE6ks8AAAAEXN-oqQA","startCursor":null,"endCursor":null}},"title":"Activity · pgbouncer/pgbouncer"}