{"payload":{"feedbackUrl":"https://github.com/orgs/community/discussions/53140","repo":{"id":309512982,"defaultBranch":"dev","name":"redpanda","ownerLogin":"redpanda-data","currentUserCanPush":false,"isFork":false,"isEmpty":false,"createdAt":"2020-11-02T22:43:36.000Z","ownerAvatar":"https://avatars.githubusercontent.com/u/49406389?v=4","public":true,"private":false,"isOrgOwned":true},"refInfo":{"name":"","listCacheKey":"v0:1716597680.0","currentOid":""},"activityList":{"items":[{"before":null,"after":"e248840863bae3a32f8cfda3d78577c39a9df810","ref":"refs/heads/jackie-bethos-gitignore-2","pushedAt":"2024-05-25T00:41:20.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"jackietung-redpanda","name":null,"path":"/jackietung-redpanda","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/166445566?s=80&v=4"},"commit":{"message":"[v24.1.x] ignore benthos local repo dir","shortMessageHtmlLink":"[v24.1.x] ignore benthos local repo dir"}},{"before":null,"after":"e248840863bae3a32f8cfda3d78577c39a9df810","ref":"refs/heads/jackie-bethos-gitignore","pushedAt":"2024-05-24T23:55:09.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"jackietung-redpanda","name":null,"path":"/jackietung-redpanda","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/166445566?s=80&v=4"},"commit":{"message":"[v24.1.x] ignore benthos local repo dir","shortMessageHtmlLink":"[v24.1.x] ignore benthos local repo dir"}},{"before":null,"after":"e517ce3806821a7be14f3fccba9d377c4b481c53","ref":"refs/heads/jackie-rpk-connect-shim","pushedAt":"2024-05-24T23:30:02.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"jackietung-redpanda","name":null,"path":"/jackietung-redpanda","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/166445566?s=80&v=4"},"commit":{"message":"wip","shortMessageHtmlLink":"wip"}},{"before":"54708d3d4f76770582ac53bb4ca451557748f23e","after":null,"ref":"refs/heads/feat/core-2069/fips-s3-virtual-address","pushedAt":"2024-05-24T21:58:30.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"andijcr","name":"Andrea Barbadoro","path":"/andijcr","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/304799?s=80&v=4"}},{"before":"361aeb2b1d9760a20b46148289c489b8d28d043a","after":"bd8602b08078c96a30487efa79089b4b096520ff","ref":"refs/heads/dev","pushedAt":"2024-05-24T21:29:09.000Z","pushType":"pr_merge","commitsCount":3,"pusher":{"login":"dotnwat","name":"Noah Watkins","path":"/dotnwat","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/242417?s=80&v=4"},"commit":{"message":"Merge pull request #18651 from dotnwat/x\n\nconfig: break dependencies on schema registry enums","shortMessageHtmlLink":"Merge pull request #18651 from dotnwat/x"}},{"before":"3e554cd51f030a0ec269c0fdf32b260024958fab","after":"959975522a3425bd2f2d0960805123f30e265d61","ref":"refs/heads/v23.3.x","pushedAt":"2024-05-24T21:17:01.000Z","pushType":"pr_merge","commitsCount":2,"pusher":{"login":"r-vasquez","name":"Rogger Vasquez","path":"/r-vasquez","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/59714880?s=80&v=4"},"commit":{"message":"Merge pull request #18646 from vbotbuildovich/backport-pr-18221-v23.3.x-929\n\n[v23.3.x] rpk: topic describe supports --regex flag","shortMessageHtmlLink":"Merge pull request #18646 from vbotbuildovich/backport-pr-18221-v23.3…"}},{"before":"69d8579e2c1aacb0d5963b74a6eec79888e155ba","after":null,"ref":"refs/heads/PESDLC-876-enable-boto3-logging","pushedAt":"2024-05-24T21:03:43.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"savex","name":"Alex","path":"/savex","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1675155?s=80&v=4"}},{"before":"ff71a747e31275646a88f716bf9fca54cea2a8bf","after":"361aeb2b1d9760a20b46148289c489b8d28d043a","ref":"refs/heads/dev","pushedAt":"2024-05-24T21:03:42.000Z","pushType":"pr_merge","commitsCount":3,"pusher":{"login":"savex","name":"Alex","path":"/savex","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1675155?s=80&v=4"},"commit":{"message":"Merge pull request #18491 from redpanda-data/PESDLC-876-enable-boto3-logging\n\nPESDLC-876 Add Filehandlers to boto3 loggers","shortMessageHtmlLink":"Merge pull request #18491 from redpanda-data/PESDLC-876-enable-boto3-…"}},{"before":"122097b331d3267ac70e73ef8768b8b5fac76e85","after":"ff71a747e31275646a88f716bf9fca54cea2a8bf","ref":"refs/heads/dev","pushedAt":"2024-05-24T11:14:00.000Z","pushType":"pr_merge","commitsCount":2,"pusher":{"login":"nvartolomei","name":"Nicolae Vartolomei","path":"/nvartolomei","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/543193?s=80&v=4"},"commit":{"message":"Merge pull request #18508 from nvartolomei/nv/saturating-kafka-next-offset\n\nmodel: make kafka::next_offset saturate rather than overflow","shortMessageHtmlLink":"Merge pull request #18508 from nvartolomei/nv/saturating-kafka-next-o…"}},{"before":"90878d7ad3deee487e8c9cb7dbb91a94e2fd2842","after":"498174cc9e824a98738ddac055ace50af15604d7","ref":"refs/heads/stephan/truncate-plus-allocate","pushedAt":"2024-05-24T09:41:33.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"StephanDollberg","name":"Stephan Dollberg","path":"/StephanDollberg","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/2904723?s=80&v=4"},"commit":{"message":"segment_appender: truncate in addition to allocate\n\nIn the segment_appender we adaptively increase the file size of log\nsegments using `ss::file::allocate`.\n\nThis however doesn't immediately do what's intended. Internally it calls\n`fallocate` with the `FALLOC_FL_KEEP_SIZE|FALLOC_FL_ZERO_RANGE` flags.\nThe former leads to the logical file size not being extended.\n\nExtending the logical file size however seems to have been the intention\nof 7d7706778918772336dda3a5cc7f63b665f106bf. We do already handle zero\nbytes at the end fine either way as we are writing zero appended 4k\nchunks anyway.\n\nBesides this side effect there is other reasons why we want to update\nthe logical file size immediately.\n\nFirst, this means that on every write we will need to update the file\nsize which makes fsync more expensive.\n\nSecond, seastar considers XFS an append-challenged file system. This\nmeans that seastar will avoid having size changing and non-size changing\noperations outstanding at the same time. They will be queued internally\nin the ioqueue. Because of how `allocate` works above all our writes\nwill always be considered as \"appending\" as we never updated the logical\nfile size.\n\nTo optimize the queued operations seastar employs certain\n\"optimizations\" in `append_challenged_posix_file_impl::optimize_queue`.\nBecause all our operations are appending this causes a continuous stream\nof implicit `ftruncate` syscalls of about 100/s per shard.\n\n```\nroot:/tmp# perf trace -t 11990 -s -e ftruncate -- sleep 5\n\n Summary of events:\n\n redpanda (11990), 1122 events, 100.0%\n\n syscall calls errors total min avg max stddev\n (msec) (msec) (msec) (msec) (%)\n --------------- -------- ------ -------- --------- --------- --------- ------\n ftruncate 561 0 32.056 0.013 0.057 0.265 1.85%\n```\n\nTo avoid the two aforementioned issues this patch adds an explicit\n`truncate` after our `allocate` call. Note that we still need both as\n`ftruncate` alone doesn't preallocate blocks so it's a lot less\nperformant on its own.\n\nIn a medium IOPS OMB workload we see a drop in p99 producer latency from\n~14ms to ~10ms. Threaded fallbacks and hence steal time are down by\nabout 5 and 10% respectively.\n\nFurther we can also see the `fsync` times distribution slightly\nimproved:\n\nBefore:\n\n```\nroot:/tmp# xfsdist-bpfcc 5 1\nTracing XFS operation latency... Hit Ctrl-C to end.\n\n14:07:37:\n\noperation = b'fsync'\n usecs : count distribution\n 0 -> 1 : 0 | |\n 2 -> 3 : 55 | |\n 4 -> 7 : 11872 |***** |\n 8 -> 15 : 38206 |***************** |\n 16 -> 31 : 18405 |******** |\n 32 -> 63 : 11474 |***** |\n 64 -> 127 : 29074 |************* |\n 128 -> 255 : 88607 |****************************************|\n 256 -> 511 : 27403 |************ |\n 512 -> 1023 : 557 | |\n 1024 -> 2047 : 28 | |\n 2048 -> 4095 : 23 |\n```\n\nAfter:\n\n```\nroot:/tmp# xfsdist-bpfcc 5 1\nTracing XFS operation latency... Hit Ctrl-C to end.\n\n13:57:45:\n\noperation = b'fsync'\n usecs : count distribution\n 0 -> 1 : 0 | |\n 2 -> 3 : 56 | |\n 4 -> 7 : 15472 |******* |\n 8 -> 15 : 45440 |*********************** |\n 16 -> 31 : 22825 |*********** |\n 32 -> 63 : 13378 |****** |\n 64 -> 127 : 31979 |**************** |\n 128 -> 255 : 77529 |****************************************|\n 256 -> 511 : 18116 |********* |\n 512 -> 1023 : 250 | |\n 1024 -> 2047 : 1 |\n```\n\nFurther we can also observe the logial file size adapting during a run:\n\nBefore:\n\n```\nroot:/tmp# for i in {0..5} ; do ll /var/lib/redpanda/data/kafka/test-topic-zUv2_Hs-0000/0_32/ | grep log ; sleep 1 ; done\n-rw-r--r-- 1 redpanda redpanda 37502976 May 21 10:10 0-1-v1.log\n-rw-r--r-- 1 redpanda redpanda 39337984 May 21 10:10 0-1-v1.log\n-rw-r--r-- 1 redpanda redpanda 41267200 May 21 10:10 0-1-v1.log\n-rw-r--r-- 1 redpanda redpanda 43479040 May 21 10:10 0-1-v1.log\n-rw-r--r-- 1 redpanda redpanda 45416448 May 21 10:10 0-1-v1.log\n-rw-r--r-- 1 redpanda redpanda 47005696 May 21 10:10 0-1-v1.log\n```\n\nAfter:\n\n```\nroot:/tmp# for i in {0..5} ; do ll /var/lib/redpanda/data/kafka/test-topic-Yk-n3mY-0000/0_35/ | grep log ; sleep 1 ; done\n-rw-r--r-- 1 redpanda redpanda 67108864 May 21 10:22 0-1-v1.log\n-rw-r--r-- 1 redpanda redpanda 67108864 May 21 10:22 0-1-v1.log\n-rw-r--r-- 1 redpanda redpanda 67108864 May 21 10:22 0-1-v1.log\n-rw-r--r-- 1 redpanda redpanda 67108864 May 21 10:22 0-1-v1.log\n-rw-r--r-- 1 redpanda redpanda 67108864 May 21 10:22 0-1-v1.log\n-rw-r--r-- 1 redpanda redpanda 67108864 May 21 10:22 0-1-v1.log\n```\n\nNote that longterm we will want to improve/change seastar to allow doing\na \"full\" fallocate but just adding the truncate is the easier change for\nnow.","shortMessageHtmlLink":"segment_appender: truncate in addition to allocate"}},{"before":"90de5a23b0b0f1b9fa654c4171efd8d0d8d2f216","after":"721cd6ea05a257ffae7d3fc0be6fd59329e51da1","ref":"refs/heads/v24.1.x","pushedAt":"2024-05-24T08:42:16.000Z","pushType":"pr_merge","commitsCount":6,"pusher":{"login":"andijcr","name":"Andrea Barbadoro","path":"/andijcr","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/304799?s=80&v=4"},"commit":{"message":"Merge pull request #18437 from vbotbuildovich/backport-pr-18397-v24.1.x-333\n\n[v24.1.x] [CORE-96]: admin/server fix set_log_level handling of overlapping expirations","shortMessageHtmlLink":"Merge pull request #18437 from vbotbuildovich/backport-pr-18397-v24.1…"}},{"before":"0a71009813f808d5a6bbeaea187a4b90badf43b3","after":"90878d7ad3deee487e8c9cb7dbb91a94e2fd2842","ref":"refs/heads/stephan/truncate-plus-allocate","pushedAt":"2024-05-24T08:14:12.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"StephanDollberg","name":"Stephan Dollberg","path":"/StephanDollberg","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/2904723?s=80&v=4"},"commit":{"message":"segment_appender: truncate in addition to allocate\n\nIn the segment_appender we adaptively increase the file size of log\nsegments using `ss::file::allocate`.\n\nThis however doesn't immediately do what's intended. Internally it calls\n`fallocate` with the `FALLOC_FL_KEEP_SIZE|FALLOC_FL_ZERO_RANGE` flags.\nThe former leads to the logical file size not being extended.\n\nExtending the logical file size however seems to have been the intention\nof 7d7706778918772336dda3a5cc7f63b665f106bf. We do already handle zero\nbytes at the end fine either way as we are writing zero appended 4k\nchunks anyway.\n\nBesides this side effect there is other reasons why we want to update\nthe logical file size immediately.\n\nFirst, this means that on every write we will need to update the file\nsize which makes fsync more expensive.\n\nSecond, seastar considers XFS an append-challenged file system. This\nmeans that seastar will avoid having size changing and non-size changing\noperations outstanding at the same time. They will be queued internally\nin the ioqueue. Because of how `allocate` works above all our writes\nwill always be considered as \"appending\" as we never updated the logical\nfile size.\n\nTo optimize the queued operations seastar employs certain\n\"optimizations\" in `append_challenged_posix_file_impl::optimize_queue`.\nBecause all our operations are appending this causes a continuous stream\nof implicit `ftruncate` syscalls of about 100/s per shard.\n\n```\nroot:/tmp# perf trace -t 11990 -s -e ftruncate -- sleep 5\n\n Summary of events:\n\n redpanda (11990), 1122 events, 100.0%\n\n syscall calls errors total min avg max stddev\n (msec) (msec) (msec) (msec) (%)\n --------------- -------- ------ -------- --------- --------- --------- ------\n ftruncate 561 0 32.056 0.013 0.057 0.265 1.85%\n```\n\nTo avoid the two aforementioned issues this patch adds an explicit\n`truncate` after our `allocate` call. Note that we still need both as\n`ftruncate` alone doesn't preallocate blocks so it's a lot less\nperformant on its own.\n\nIn a medium IOPS OMB workload we see a drop in p99 producer latency from\n~14ms to ~10ms. Threaded fallbacks and hence steal time are down by\nabout 5 and 10% respectively.\n\nFurther we can also see the `fsync` times distribution slightly\nimproved:\n\nBefore:\n\n```\nroot:/tmp# xfsdist-bpfcc 5 1\nTracing XFS operation latency... Hit Ctrl-C to end.\n\n14:07:37:\n\noperation = b'fsync'\n usecs : count distribution\n 0 -> 1 : 0 | |\n 2 -> 3 : 55 | |\n 4 -> 7 : 11872 |***** |\n 8 -> 15 : 38206 |***************** |\n 16 -> 31 : 18405 |******** |\n 32 -> 63 : 11474 |***** |\n 64 -> 127 : 29074 |************* |\n 128 -> 255 : 88607 |****************************************|\n 256 -> 511 : 27403 |************ |\n 512 -> 1023 : 557 | |\n 1024 -> 2047 : 28 | |\n 2048 -> 4095 : 23 |\n```\n\nAfter:\n\n```\nroot:/tmp# xfsdist-bpfcc 5 1\nTracing XFS operation latency... Hit Ctrl-C to end.\n\n13:57:45:\n\noperation = b'fsync'\n usecs : count distribution\n 0 -> 1 : 0 | |\n 2 -> 3 : 56 | |\n 4 -> 7 : 15472 |******* |\n 8 -> 15 : 45440 |*********************** |\n 16 -> 31 : 22825 |*********** |\n 32 -> 63 : 13378 |****** |\n 64 -> 127 : 31979 |**************** |\n 128 -> 255 : 77529 |****************************************|\n 256 -> 511 : 18116 |********* |\n 512 -> 1023 : 250 | |\n 1024 -> 2047 : 1 |\n```\n\nFurther we can also observe the logial file size adapting during a run:\n\nBefore:\n\n```\nroot:/tmp# for i in {0..5} ; do ll /var/lib/redpanda/data/kafka/test-topic-zUv2_Hs-0000/0_32/ | grep log ; sleep 1 ; done\n-rw-r--r-- 1 redpanda redpanda 37502976 May 21 10:10 0-1-v1.log\n-rw-r--r-- 1 redpanda redpanda 39337984 May 21 10:10 0-1-v1.log\n-rw-r--r-- 1 redpanda redpanda 41267200 May 21 10:10 0-1-v1.log\n-rw-r--r-- 1 redpanda redpanda 43479040 May 21 10:10 0-1-v1.log\n-rw-r--r-- 1 redpanda redpanda 45416448 May 21 10:10 0-1-v1.log\n-rw-r--r-- 1 redpanda redpanda 47005696 May 21 10:10 0-1-v1.log\n```\n\nAfter:\n\n```\nroot:/tmp# for i in {0..5} ; do ll /var/lib/redpanda/data/kafka/test-topic-Yk-n3mY-0000/0_35/ | grep log ; sleep 1 ; done\n-rw-r--r-- 1 redpanda redpanda 67108864 May 21 10:22 0-1-v1.log\n-rw-r--r-- 1 redpanda redpanda 67108864 May 21 10:22 0-1-v1.log\n-rw-r--r-- 1 redpanda redpanda 67108864 May 21 10:22 0-1-v1.log\n-rw-r--r-- 1 redpanda redpanda 67108864 May 21 10:22 0-1-v1.log\n-rw-r--r-- 1 redpanda redpanda 67108864 May 21 10:22 0-1-v1.log\n-rw-r--r-- 1 redpanda redpanda 67108864 May 21 10:22 0-1-v1.log\n```\n\nNote that longterm we will want to improve/change seastar to allow doing\na \"full\" fallocate but just adding the truncate is the easier change for\nnow.","shortMessageHtmlLink":"segment_appender: truncate in addition to allocate"}},{"before":"7d23744fccb60450061661499d472e63c541e4d5","after":"122097b331d3267ac70e73ef8768b8b5fac76e85","ref":"refs/heads/dev","pushedAt":"2024-05-24T05:46:19.000Z","pushType":"pr_merge","commitsCount":3,"pusher":{"login":"andrwng","name":"Andrew Wong","path":"/andrwng","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/20144987?s=80&v=4"},"commit":{"message":"Merge pull request #18521 from andrwng/mvlog-segment-rolls\n\nmvlog: log basics for segment rolls","shortMessageHtmlLink":"Merge pull request #18521 from andrwng/mvlog-segment-rolls"}},{"before":"0208f20d06114db3a19a6137eed710a366a9442f","after":"7d23744fccb60450061661499d472e63c541e4d5","ref":"refs/heads/dev","pushedAt":"2024-05-24T01:19:44.000Z","pushType":"pr_merge","commitsCount":2,"pusher":{"login":"michael-redpanda","name":"Mike Boquard","path":"/michael-redpanda","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/122312462?s=80&v=4"},"commit":{"message":"Merge pull request #18629 from michael-redpanda/CORE-3066-Fix-fips-mode\n\napp/crypto: Fixed issue with RP not going into FIPS mode","shortMessageHtmlLink":"Merge pull request #18629 from michael-redpanda/CORE-3066-Fix-fips-mode"}},{"before":"138faa1cfe9bcf5adcb98ac4066cfaa658d6e9a0","after":"0208f20d06114db3a19a6137eed710a366a9442f","ref":"refs/heads/dev","pushedAt":"2024-05-24T00:40:47.000Z","pushType":"pr_merge","commitsCount":3,"pusher":{"login":"dotnwat","name":"Noah Watkins","path":"/dotnwat","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/242417?s=80&v=4"},"commit":{"message":"Merge pull request #18642 from dotnwat/break-config-cloud-storage-client-dep\n\nconfig: break dependency on cloud_storage_clients","shortMessageHtmlLink":"Merge pull request #18642 from dotnwat/break-config-cloud-storage-cli…"}},{"before":"165b95294c328bf31e4176db5329773ab681333e","after":"3e554cd51f030a0ec269c0fdf32b260024958fab","ref":"refs/heads/v23.3.x","pushedAt":"2024-05-23T20:02:10.000Z","pushType":"pr_merge","commitsCount":5,"pusher":{"login":"r-vasquez","name":"Rogger Vasquez","path":"/r-vasquez","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/59714880?s=80&v=4"},"commit":{"message":"Merge pull request #18634 from vbotbuildovich/backport-pr-18628-v23.3.x-635\n\n[v23.3.x] rpk: small fixes.","shortMessageHtmlLink":"Merge pull request #18634 from vbotbuildovich/backport-pr-18628-v23.3…"}},{"before":"d907f6fc621403d2a983033060ed13bda0cd40c1","after":"138faa1cfe9bcf5adcb98ac4066cfaa658d6e9a0","ref":"refs/heads/dev","pushedAt":"2024-05-23T18:56:45.000Z","pushType":"pr_merge","commitsCount":2,"pusher":{"login":"r-vasquez","name":"Rogger Vasquez","path":"/r-vasquez","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/59714880?s=80&v=4"},"commit":{"message":"Merge pull request #18221 from daisukebe/rpk-describe-all-topics\n\nrpk: topic describe supports --regex flag","shortMessageHtmlLink":"Merge pull request #18221 from daisukebe/rpk-describe-all-topics"}},{"before":"487647852c75b2cd32db0e6fafab6de56331c65c","after":"0a71009813f808d5a6bbeaea187a4b90badf43b3","ref":"refs/heads/stephan/truncate-plus-allocate","pushedAt":"2024-05-23T17:54:16.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"StephanDollberg","name":"Stephan Dollberg","path":"/StephanDollberg","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/2904723?s=80&v=4"},"commit":{"message":"rpk: Increase default controller log size limit\n\nOtherwise the controller log would now be skipped because on the average\nsystem where the default falloc step is 32MB it would now already be too\nbig for the default 20MB limit.\n\nNote that this doesn't actually bloat the size of the debug bundle as\nany zero padding basically compresses perfectly anyway.\n\n20 MB also feels kinda low for something important as the controller\nlog.","shortMessageHtmlLink":"rpk: Increase default controller log size limit"}},{"before":"8271f5197ad4714d70eddb6f58f12c2ae67cb068","after":"487647852c75b2cd32db0e6fafab6de56331c65c","ref":"refs/heads/stephan/truncate-plus-allocate","pushedAt":"2024-05-23T17:51:57.000Z","pushType":"push","commitsCount":3,"pusher":{"login":"StephanDollberg","name":"Stephan Dollberg","path":"/StephanDollberg","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/2904723?s=80&v=4"},"commit":{"message":"rpk: Increase default controller log size limit\n\nOtherwise the controller log would now be skipped because on the average\nsystem where the default falloc step is 32MB it would now already be too\nbig for the default 20MB limit.\n\nNote that this doesn't actually bloat the size of the debug bundle as\nany zero padding basically compresses perfectly anyway.\n\n20 MB also feels kinda low for something important as the controller\nlog.","shortMessageHtmlLink":"rpk: Increase default controller log size limit"}},{"before":"9cf6bdccd0cbce3b1cb235f3574d6326e22ae07d","after":"8271f5197ad4714d70eddb6f58f12c2ae67cb068","ref":"refs/heads/stephan/truncate-plus-allocate","pushedAt":"2024-05-23T13:07:30.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"StephanDollberg","name":"Stephan Dollberg","path":"/StephanDollberg","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/2904723?s=80&v=4"},"commit":{"message":"use compute_storage for data_checksum","shortMessageHtmlLink":"use compute_storage for data_checksum"}},{"before":"594c3eda5f025c3cd5810cef0980b965bc5b447e","after":"d907f6fc621403d2a983033060ed13bda0cd40c1","ref":"refs/heads/dev","pushedAt":"2024-05-23T12:38:58.000Z","pushType":"pr_merge","commitsCount":8,"pusher":{"login":"WillemKauf","name":"Willem Kaufmann","path":"/WillemKauf","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7502731?s=80&v=4"},"commit":{"message":"Merge pull request #18106 from WillemKauf/path_style_url_testing\n\nCORE-2154: cloud_storage: path-style verification tests","shortMessageHtmlLink":"Merge pull request #18106 from WillemKauf/path_style_url_testing"}},{"before":"1cae54a38e0e27883d3cff1224a90b64167d5e4b","after":"90de5a23b0b0f1b9fa654c4171efd8d0d8d2f216","ref":"refs/heads/v24.1.x","pushedAt":"2024-05-23T12:21:57.000Z","pushType":"pr_merge","commitsCount":2,"pusher":{"login":"BenPope","name":"Ben Pope","path":"/BenPope","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/922500?s=80&v=4"},"commit":{"message":"Merge pull request #18639 from vbotbuildovich/backport-pr-18635-v24.1.x-775\n\n[v24.1.x] schema_registry: Make mode_mutability: true by default","shortMessageHtmlLink":"Merge pull request #18639 from vbotbuildovich/backport-pr-18635-v24.1…"}},{"before":"d474332b5e65ceae3bf6b995fb29bc0061359008","after":"9cf6bdccd0cbce3b1cb235f3574d6326e22ae07d","ref":"refs/heads/stephan/truncate-plus-allocate","pushedAt":"2024-05-23T09:51:43.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"StephanDollberg","name":"Stephan Dollberg","path":"/StephanDollberg","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/2904723?s=80&v=4"},"commit":{"message":"use compute_storage for data_checksum","shortMessageHtmlLink":"use compute_storage for data_checksum"}},{"before":"3b6022e9db3955fa581cbcc1d108113b656f07ed","after":"594c3eda5f025c3cd5810cef0980b965bc5b447e","ref":"refs/heads/dev","pushedAt":"2024-05-23T07:41:58.000Z","pushType":"pr_merge","commitsCount":2,"pusher":{"login":"BenPope","name":"Ben Pope","path":"/BenPope","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/922500?s=80&v=4"},"commit":{"message":"Merge pull request #18635 from BenPope/schema-registry-mode-mutability-default\n\nschema_registry: Make mode_mutability: true by default","shortMessageHtmlLink":"Merge pull request #18635 from BenPope/schema-registry-mode-mutabilit…"}},{"before":"2e3e74609a8904bfa4527c354c978f7a0d7a4102","after":"165b95294c328bf31e4176db5329773ab681333e","ref":"refs/heads/v23.3.x","pushedAt":"2024-05-23T06:10:52.000Z","pushType":"pr_merge","commitsCount":6,"pusher":{"login":"piyushredpanda","name":null,"path":"/piyushredpanda","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/102703159?s=80&v=4"},"commit":{"message":"Merge pull request #18573 from bharathv/snapshot_producer_fixes\n\nrm_stm: couple of stability fixes noticed when down scaling max_concurrent_producer_ids","shortMessageHtmlLink":"Merge pull request #18573 from bharathv/snapshot_producer_fixes"}},{"before":"7506261a21a0098c29cec49a7a4e6a37da6be97a","after":"1cae54a38e0e27883d3cff1224a90b64167d5e4b","ref":"refs/heads/v24.1.x","pushedAt":"2024-05-23T05:49:29.000Z","pushType":"pr_merge","commitsCount":5,"pusher":{"login":"piyushredpanda","name":null,"path":"/piyushredpanda","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/102703159?s=80&v=4"},"commit":{"message":"Merge pull request #18632 from vbotbuildovich/backport-pr-18628-v24.1.x-54\n\n[v24.1.x] rpk: small fixes.","shortMessageHtmlLink":"Merge pull request #18632 from vbotbuildovich/backport-pr-18628-v24.1…"}},{"before":"607753cb8a5a8d90505b239627cee63f45536a9f","after":"7506261a21a0098c29cec49a7a4e6a37da6be97a","ref":"refs/heads/v24.1.x","pushedAt":"2024-05-23T05:46:11.000Z","pushType":"pr_merge","commitsCount":2,"pusher":{"login":"piyushredpanda","name":null,"path":"/piyushredpanda","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/102703159?s=80&v=4"},"commit":{"message":"Merge pull request #18428 from vbotbuildovich/backport-pr-18304-v24.1.x-168\n\n[v24.1.x] http: Fix double call to stop() in http::client","shortMessageHtmlLink":"Merge pull request #18428 from vbotbuildovich/backport-pr-18304-v24.1…"}},{"before":"6323b8dac141854c341c1cd156b6fb9b8fef5da0","after":"607753cb8a5a8d90505b239627cee63f45536a9f","ref":"refs/heads/v24.1.x","pushedAt":"2024-05-23T05:44:03.000Z","pushType":"pr_merge","commitsCount":2,"pusher":{"login":"piyushredpanda","name":null,"path":"/piyushredpanda","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/102703159?s=80&v=4"},"commit":{"message":"Merge pull request #18440 from vbotbuildovich/backport-pr-18303-v24.1.x-523\n\n[v24.1.x] [CORE-2814]: utils/file_io: fix double closing of ss::file in write_fully","shortMessageHtmlLink":"Merge pull request #18440 from vbotbuildovich/backport-pr-18303-v24.1…"}},{"before":"5ff21da71201a94c07fc2812a8ada4befb2333eb","after":"6323b8dac141854c341c1cd156b6fb9b8fef5da0","ref":"refs/heads/v24.1.x","pushedAt":"2024-05-23T05:43:52.000Z","pushType":"pr_merge","commitsCount":2,"pusher":{"login":"piyushredpanda","name":null,"path":"/piyushredpanda","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/102703159?s=80&v=4"},"commit":{"message":"Merge pull request #18459 from vbotbuildovich/backport-pr-18026-v24.1.x-126\n\n[v24.1.x] rpk: add transfer-leadership command","shortMessageHtmlLink":"Merge pull request #18459 from vbotbuildovich/backport-pr-18026-v24.1…"}},{"before":"271c5334d55ed9683a50963903a3b0113094351b","after":"5ff21da71201a94c07fc2812a8ada4befb2333eb","ref":"refs/heads/v24.1.x","pushedAt":"2024-05-23T05:43:37.000Z","pushType":"pr_merge","commitsCount":6,"pusher":{"login":"piyushredpanda","name":null,"path":"/piyushredpanda","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/102703159?s=80&v=4"},"commit":{"message":"Merge pull request #18638 from vbotbuildovich/backport-pr-18626-v24.1.x-939\n\n[v24.1.x] Fixed `state_machine_manager` stall","shortMessageHtmlLink":"Merge pull request #18638 from vbotbuildovich/backport-pr-18626-v24.1…"}}],"hasNextPage":true,"hasPreviousPage":false,"activityType":"all","actor":null,"timePeriod":"all","sort":"DESC","perPage":30,"cursor":"djE6ks8AAAAEU301pAA","startCursor":null,"endCursor":null}},"title":"Activity · redpanda-data/redpanda"}