{"payload":{"feedbackUrl":"https://github.com/orgs/community/discussions/53140","repo":{"id":182148991,"defaultBranch":"main","name":"ovn","ownerLogin":"ovn-org","currentUserCanPush":false,"isFork":false,"isEmpty":false,"createdAt":"2019-04-18T19:54:33.000Z","ownerAvatar":"https://avatars.githubusercontent.com/u/33884263?v=4","public":true,"private":false,"isOrgOwned":true},"refInfo":{"name":"","listCacheKey":"v0:1715362436.0","currentOid":""},"activityList":{"items":[{"before":"7b65a65402963caf1cf7ff1b2152290f02277750","after":"ff11ab416c230ee36c18b52d86a505a27cb932c8","ref":"refs/heads/branch-24.03","pushedAt":"2024-05-21T20:40:35.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"hzhou8","name":"Han Zhou","path":"/hzhou8","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7600054?s=80&v=4"},"commit":{"message":"northd: Fix an issue wrt mac binding aging.\n\nIssue:\nIn case of a Logical_Router without mac_binding_age_threshold set or a\nLogical_Router with an incorrectly formatted mac_binding_threshold\noption, entries were not being purged from the Mac Binding table in\nSouthBound.\n\nThis was because in the function `en_mac_binding_aging_run` in case of\nan invalid mac_binding_threshold entry or if mac_binding_threshold is\nnot set we are returning from the loop instead of iterating through the\nremaining LRs. As a result, subsequent runs of the aging_waker node are\nalso not scehduled and we end up not purging any MAC Bindings.\n\nFix:\nThis patch fixes this issue by changing the return to a continue so that\nwe skip the current LR but continue processing for the remaining LRs.\n\nFixes: 78851b6ffb58 (\"Support CIDR-based MAC binding aging threshold.\")\nSigned-off-by: Indrajitt Valsaraj \nAcked-by: Naveen Yerramneni \nAcked-by: Ales Musil \nSigned-off-by: Han Zhou \n(cherry picked from commit 7abae8142841ba3083e08678f5e5d01039dcc428)","shortMessageHtmlLink":"northd: Fix an issue wrt mac binding aging."}},{"before":"47915c4c517c634dec919cfd60295db0d0bedfa7","after":"7abae8142841ba3083e08678f5e5d01039dcc428","ref":"refs/heads/main","pushedAt":"2024-05-21T20:11:24.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"hzhou8","name":"Han Zhou","path":"/hzhou8","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7600054?s=80&v=4"},"commit":{"message":"northd: Fix an issue wrt mac binding aging.\n\nIssue:\nIn case of a Logical_Router without mac_binding_age_threshold set or a\nLogical_Router with an incorrectly formatted mac_binding_threshold\noption, entries were not being purged from the Mac Binding table in\nSouthBound.\n\nThis was because in the function `en_mac_binding_aging_run` in case of\nan invalid mac_binding_threshold entry or if mac_binding_threshold is\nnot set we are returning from the loop instead of iterating through the\nremaining LRs. As a result, subsequent runs of the aging_waker node are\nalso not scehduled and we end up not purging any MAC Bindings.\n\nFix:\nThis patch fixes this issue by changing the return to a continue so that\nwe skip the current LR but continue processing for the remaining LRs.\n\nFixes: 78851b6ffb58 (\"Support CIDR-based MAC binding aging threshold.\")\nSigned-off-by: Indrajitt Valsaraj \nAcked-by: Naveen Yerramneni \nAcked-by: Ales Musil \nSigned-off-by: Han Zhou ","shortMessageHtmlLink":"northd: Fix an issue wrt mac binding aging."}},{"before":"0440a083d2bb024fe3a8263f72c7c749cc9fae5a","after":"7b65a65402963caf1cf7ff1b2152290f02277750","ref":"refs/heads/branch-24.03","pushedAt":"2024-05-21T13:33:04.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"ci: Pin Fedora version for the build-rpm job.\n\nKeep the Fedora version pinned to 40 for the build-rpm.\n\nFixes: 8f18b3b6c52f (\"ci: Keep the container version pinned.\")\nSigned-off-by: Ales Musil \nSigned-off-by: Numan Siddique \n(cherry picked from commit a02ac71fa0774c0e57c151353ff2b4cc7a4c7207)","shortMessageHtmlLink":"ci: Pin Fedora version for the build-rpm job."}},{"before":"e5d093cc243e1ad478eaeddad445523a56f2f676","after":"47915c4c517c634dec919cfd60295db0d0bedfa7","ref":"refs/heads/main","pushedAt":"2024-05-16T15:56:02.000Z","pushType":"push","commitsCount":7,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"ci: Bump the Fedora container to 40.\n\nNow that all failures were resolved for Fedora 40\nwe can bump the version in the container.\n\nSigned-off-by: Ales Musil \nSigned-off-by: Numan Siddique ","shortMessageHtmlLink":"ci: Bump the Fedora container to 40."}},{"before":"5507cc4653443bfdf6d047334d2bee8a0499acc0","after":"45e4940e7503dcd922aaefd2e58bdd61399f6cf4","ref":"refs/heads/branch-23.03","pushedAt":"2024-05-16T15:53:21.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"controller: Avoid use after free in LB I-P.\n\nAvoid use after free in scenario when controller received LB deletion\nafter the DB was reconnected. The reconnect led to idl clearing up\nthe \"old\" structs, one of them being the LB. However, during recompute\nthe struct was referenced when it was already gone.\n\nClear the whole objdep_mgr instead of going one-by-one during recompute.\n\n==143949==ERROR: AddressSanitizer: heap-use-after-free\nREAD of size 4 at 0x5130000280d0 thread T0\n 0 0x61c3c9 in lb_data_local_lb_remove controller/ovn-controller.c:2978:5\n 1 0x5fd4df in en_lb_data_run controller/ovn-controller.c:3063:9\n 2 0x6fe0d9 in engine_recompute lib/inc-proc-eng.c:415:5\n 3 0x6fbdc2 in engine_run_node lib/inc-proc-eng.c:477:9\n 4 0x6fbdc2 in engine_run lib/inc-proc-eng.c:528:9\n 5 0x5f39a0 in main controller/ovn-controller.c\n\nFixes: 8382127186bf (\"controller: Store load balancer data in separate node\")\nReported-at: https://issues.redhat.com/browse/FDP-610\nSigned-off-by: Ales Musil \nSigned-off-by: Numan Siddique \n(cherry picked from commit e5d093cc243e1ad478eaeddad445523a56f2f676)","shortMessageHtmlLink":"controller: Avoid use after free in LB I-P."}},{"before":"a92e4bb38c54bba5e3c9eaf312b19eb86484995f","after":"f1dcfeef80b0f55d87d778df6f0b219a19cd70ef","ref":"refs/heads/branch-23.06","pushedAt":"2024-05-16T15:52:57.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"controller: Avoid use after free in LB I-P.\n\nAvoid use after free in scenario when controller received LB deletion\nafter the DB was reconnected. The reconnect led to idl clearing up\nthe \"old\" structs, one of them being the LB. However, during recompute\nthe struct was referenced when it was already gone.\n\nClear the whole objdep_mgr instead of going one-by-one during recompute.\n\n==143949==ERROR: AddressSanitizer: heap-use-after-free\nREAD of size 4 at 0x5130000280d0 thread T0\n 0 0x61c3c9 in lb_data_local_lb_remove controller/ovn-controller.c:2978:5\n 1 0x5fd4df in en_lb_data_run controller/ovn-controller.c:3063:9\n 2 0x6fe0d9 in engine_recompute lib/inc-proc-eng.c:415:5\n 3 0x6fbdc2 in engine_run_node lib/inc-proc-eng.c:477:9\n 4 0x6fbdc2 in engine_run lib/inc-proc-eng.c:528:9\n 5 0x5f39a0 in main controller/ovn-controller.c\n\nFixes: 8382127186bf (\"controller: Store load balancer data in separate node\")\nReported-at: https://issues.redhat.com/browse/FDP-610\nSigned-off-by: Ales Musil \nSigned-off-by: Numan Siddique \n(cherry picked from commit e5d093cc243e1ad478eaeddad445523a56f2f676)","shortMessageHtmlLink":"controller: Avoid use after free in LB I-P."}},{"before":"d929ea791822c502989a688a1687fbbc6950d92f","after":"a8b1a53244ef885f840be848ae99bf4e406a96ff","ref":"refs/heads/branch-23.09","pushedAt":"2024-05-16T15:52:20.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"controller: Avoid use after free in LB I-P.\n\nAvoid use after free in scenario when controller received LB deletion\nafter the DB was reconnected. The reconnect led to idl clearing up\nthe \"old\" structs, one of them being the LB. However, during recompute\nthe struct was referenced when it was already gone.\n\nClear the whole objdep_mgr instead of going one-by-one during recompute.\n\n==143949==ERROR: AddressSanitizer: heap-use-after-free\nREAD of size 4 at 0x5130000280d0 thread T0\n 0 0x61c3c9 in lb_data_local_lb_remove controller/ovn-controller.c:2978:5\n 1 0x5fd4df in en_lb_data_run controller/ovn-controller.c:3063:9\n 2 0x6fe0d9 in engine_recompute lib/inc-proc-eng.c:415:5\n 3 0x6fbdc2 in engine_run_node lib/inc-proc-eng.c:477:9\n 4 0x6fbdc2 in engine_run lib/inc-proc-eng.c:528:9\n 5 0x5f39a0 in main controller/ovn-controller.c\n\nFixes: 8382127186bf (\"controller: Store load balancer data in separate node\")\nReported-at: https://issues.redhat.com/browse/FDP-610\nSigned-off-by: Ales Musil \nSigned-off-by: Numan Siddique \n(cherry picked from commit e5d093cc243e1ad478eaeddad445523a56f2f676)","shortMessageHtmlLink":"controller: Avoid use after free in LB I-P."}},{"before":"bc2af6ab81b47cb5303efb072f56dfe3ec149649","after":"0440a083d2bb024fe3a8263f72c7c749cc9fae5a","ref":"refs/heads/branch-24.03","pushedAt":"2024-05-16T15:51:39.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"controller: Avoid use after free in LB I-P.\n\nAvoid use after free in scenario when controller received LB deletion\nafter the DB was reconnected. The reconnect led to idl clearing up\nthe \"old\" structs, one of them being the LB. However, during recompute\nthe struct was referenced when it was already gone.\n\nClear the whole objdep_mgr instead of going one-by-one during recompute.\n\n==143949==ERROR: AddressSanitizer: heap-use-after-free\nREAD of size 4 at 0x5130000280d0 thread T0\n 0 0x61c3c9 in lb_data_local_lb_remove controller/ovn-controller.c:2978:5\n 1 0x5fd4df in en_lb_data_run controller/ovn-controller.c:3063:9\n 2 0x6fe0d9 in engine_recompute lib/inc-proc-eng.c:415:5\n 3 0x6fbdc2 in engine_run_node lib/inc-proc-eng.c:477:9\n 4 0x6fbdc2 in engine_run lib/inc-proc-eng.c:528:9\n 5 0x5f39a0 in main controller/ovn-controller.c\n\nFixes: 8382127186bf (\"controller: Store load balancer data in separate node\")\nReported-at: https://issues.redhat.com/browse/FDP-610\nSigned-off-by: Ales Musil \nSigned-off-by: Numan Siddique \n(cherry picked from commit e5d093cc243e1ad478eaeddad445523a56f2f676)","shortMessageHtmlLink":"controller: Avoid use after free in LB I-P."}},{"before":"b0609e6030c13fa0dd3a1615d750c6fa6b466ec5","after":"e5d093cc243e1ad478eaeddad445523a56f2f676","ref":"refs/heads/main","pushedAt":"2024-05-15T21:28:13.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"controller: Avoid use after free in LB I-P.\n\nAvoid use after free in scenario when controller received LB deletion\nafter the DB was reconnected. The reconnect led to idl clearing up\nthe \"old\" structs, one of them being the LB. However, during recompute\nthe struct was referenced when it was already gone.\n\nClear the whole objdep_mgr instead of going one-by-one during recompute.\n\n==143949==ERROR: AddressSanitizer: heap-use-after-free\nREAD of size 4 at 0x5130000280d0 thread T0\n 0 0x61c3c9 in lb_data_local_lb_remove controller/ovn-controller.c:2978:5\n 1 0x5fd4df in en_lb_data_run controller/ovn-controller.c:3063:9\n 2 0x6fe0d9 in engine_recompute lib/inc-proc-eng.c:415:5\n 3 0x6fbdc2 in engine_run_node lib/inc-proc-eng.c:477:9\n 4 0x6fbdc2 in engine_run lib/inc-proc-eng.c:528:9\n 5 0x5f39a0 in main controller/ovn-controller.c\n\nFixes: 8382127186bf (\"controller: Store load balancer data in separate node\")\nReported-at: https://issues.redhat.com/browse/FDP-610\nSigned-off-by: Ales Musil \nSigned-off-by: Numan Siddique ","shortMessageHtmlLink":"controller: Avoid use after free in LB I-P."}},{"before":"12412b13c9e279dab4dc24abf82e076e5d3f400d","after":"b0609e6030c13fa0dd3a1615d750c6fa6b466ec5","ref":"refs/heads/main","pushedAt":"2024-05-15T21:07:50.000Z","pushType":"push","commitsCount":2,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"docs: List supported rolling upgrade paths.\n\nThe wording above is not completely clear without these scenarios\nlisted. A confused reader may incorrectly read it as:\n\n```\nOnly LTS-to-LTS is supported for rolling upgrades.\n```\n\nwhich is wrong.\n\nSigned-off-by: Ihar Hrachyshka \nAcked-by: Ales Musil \nSigned-off-by: Numan Siddique ","shortMessageHtmlLink":"docs: List supported rolling upgrade paths."}},{"before":"7f1218a52c1b3b14ab1bada344eff7bed89f368c","after":"12412b13c9e279dab4dc24abf82e076e5d3f400d","ref":"refs/heads/main","pushedAt":"2024-05-15T20:53:21.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"ovn-ctl: Support for --config-file ovsdb-server option.\n\nSince OVS 3.3.0 ovsdb-server accepts databases and remotes configuration\nvia JSON text file. This patch adds support for such option.\n\nSigned-off-by: Vladislav Odintsov \nAcked-by: Ales Musil \nSigned-off-by: Numan Siddique ","shortMessageHtmlLink":"ovn-ctl: Support for --config-file ovsdb-server option."}},{"before":"70c319085e0af16ab11de6d05d561e2dea991ce2","after":"5507cc4653443bfdf6d047334d2bee8a0499acc0","ref":"refs/heads/branch-23.03","pushedAt":"2024-05-14T21:37:46.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"northd, ic: Fix handling of ovn-appctl resume.\n\nAfter ovn-appctl resume was issued for northd or ovn-ic, there was no\nguarantee that northd or ovn-ic were waking up, potentially handling\nchanges received while they were paused..\nUsually, poll_block would be woken up by POLLHUP, but race conditions could\ncause this not to happen.\novn-controller is already properly handling the resume.\n\nThis caused the following tests to fail sporadically:\n- ovn-ic -- sync ISB status to INB\n- propagate Port_Binding.up to NB and OVS.\n\nSigned-off-by: Xavier Simonart \nAcked-by: Ales Musil \nSigned-off-by: Numan Siddique \n(cherry picked from commit 7f1218a52c1b3b14ab1bada344eff7bed89f368c)","shortMessageHtmlLink":"northd, ic: Fix handling of ovn-appctl resume."}},{"before":"05a98ef911b5435edb4d44d8f0f9ef002c0cec8f","after":"a92e4bb38c54bba5e3c9eaf312b19eb86484995f","ref":"refs/heads/branch-23.06","pushedAt":"2024-05-14T21:37:07.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"northd, ic: Fix handling of ovn-appctl resume.\n\nAfter ovn-appctl resume was issued for northd or ovn-ic, there was no\nguarantee that northd or ovn-ic were waking up, potentially handling\nchanges received while they were paused..\nUsually, poll_block would be woken up by POLLHUP, but race conditions could\ncause this not to happen.\novn-controller is already properly handling the resume.\n\nThis caused the following tests to fail sporadically:\n- ovn-ic -- sync ISB status to INB\n- propagate Port_Binding.up to NB and OVS.\n\nSigned-off-by: Xavier Simonart \nAcked-by: Ales Musil \nSigned-off-by: Numan Siddique \n(cherry picked from commit 7f1218a52c1b3b14ab1bada344eff7bed89f368c)","shortMessageHtmlLink":"northd, ic: Fix handling of ovn-appctl resume."}},{"before":"d70ff69318f4ef4553ee1da32345d5243525b698","after":"d929ea791822c502989a688a1687fbbc6950d92f","ref":"refs/heads/branch-23.09","pushedAt":"2024-05-14T21:36:37.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"northd, ic: Fix handling of ovn-appctl resume.\n\nAfter ovn-appctl resume was issued for northd or ovn-ic, there was no\nguarantee that northd or ovn-ic were waking up, potentially handling\nchanges received while they were paused..\nUsually, poll_block would be woken up by POLLHUP, but race conditions could\ncause this not to happen.\novn-controller is already properly handling the resume.\n\nThis caused the following tests to fail sporadically:\n- ovn-ic -- sync ISB status to INB\n- propagate Port_Binding.up to NB and OVS.\n\nSigned-off-by: Xavier Simonart \nAcked-by: Ales Musil \nSigned-off-by: Numan Siddique \n(cherry picked from commit 7f1218a52c1b3b14ab1bada344eff7bed89f368c)","shortMessageHtmlLink":"northd, ic: Fix handling of ovn-appctl resume."}},{"before":"9c8264e765548f7c2f2e15af217c84abc664b932","after":"bc2af6ab81b47cb5303efb072f56dfe3ec149649","ref":"refs/heads/branch-24.03","pushedAt":"2024-05-14T21:35:10.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"northd, ic: Fix handling of ovn-appctl resume.\n\nAfter ovn-appctl resume was issued for northd or ovn-ic, there was no\nguarantee that northd or ovn-ic were waking up, potentially handling\nchanges received while they were paused..\nUsually, poll_block would be woken up by POLLHUP, but race conditions could\ncause this not to happen.\novn-controller is already properly handling the resume.\n\nThis caused the following tests to fail sporadically:\n- ovn-ic -- sync ISB status to INB\n- propagate Port_Binding.up to NB and OVS.\n\nSigned-off-by: Xavier Simonart \nAcked-by: Ales Musil \nSigned-off-by: Numan Siddique \n(cherry picked from commit 7f1218a52c1b3b14ab1bada344eff7bed89f368c)","shortMessageHtmlLink":"northd, ic: Fix handling of ovn-appctl resume."}},{"before":"f56ad41703d765de10aac3bbf74707e3ab2920c4","after":"7f1218a52c1b3b14ab1bada344eff7bed89f368c","ref":"refs/heads/main","pushedAt":"2024-05-14T21:32:55.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"northd, ic: Fix handling of ovn-appctl resume.\n\nAfter ovn-appctl resume was issued for northd or ovn-ic, there was no\nguarantee that northd or ovn-ic were waking up, potentially handling\nchanges received while they were paused..\nUsually, poll_block would be woken up by POLLHUP, but race conditions could\ncause this not to happen.\novn-controller is already properly handling the resume.\n\nThis caused the following tests to fail sporadically:\n- ovn-ic -- sync ISB status to INB\n- propagate Port_Binding.up to NB and OVS.\n\nSigned-off-by: Xavier Simonart \nAcked-by: Ales Musil \nSigned-off-by: Numan Siddique ","shortMessageHtmlLink":"northd, ic: Fix handling of ovn-appctl resume."}},{"before":"2609cd9f0b6e215780b08c8ec0a12fd0f043d5ca","after":"f56ad41703d765de10aac3bbf74707e3ab2920c4","ref":"refs/heads/main","pushedAt":"2024-05-14T21:27:58.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"northd: Add lsp option force_fdb_lookup.\n\nThis option is applicable only if the lsp is of default 'type'\ni.e. type=empty_string (which is a VM (or VIF) interface) and the\nlsp also has 'unknown' addresses configured.\nIf lsp option 'force_fdb_lookup' is set to true, mac addresses\nof the lsp (if configured) are not installed in the l2 lookup\ntable of the Logical Switch pipeline. However, MAC addresses\nare learnt using the FDB Table.\n\nUsecase:\n=========\nThis option is required to reduce packet loss when VM is being\nmigrated across AZs (via VXLAN tunnel). lsp is configured in both\nAZs on different logical switches which are sharing the same IP\nsubnet. If the port has unknown address set along with MAC, IP\nthen, any packet destined to VM's MAC on destination AZ will get\nforwarded locally, which may lead to packet loss during VM migration.\n\nThis option is useful to reduce packet loss during VM migration\nby forcing the logical switch to learn MAC using FDB.\n\nThe other way to address this issue is to use pkt_clone_type\noption but this causes too much of packet duplication when there\nare multiple ports with unknown address in the logical switch.\n\nSigned-off-by: shibir-basak \nSigned-off-by: Numan Siddique ","shortMessageHtmlLink":"northd: Add lsp option force_fdb_lookup."}},{"before":"920339ec76a7800455feeb731efa18b6f5ed896b","after":"05a98ef911b5435edb4d44d8f0f9ef002c0cec8f","ref":"refs/heads/branch-23.06","pushedAt":"2024-05-10T17:33:56.000Z","pushType":"push","commitsCount":2,"pusher":{"login":"putnopvut","name":"Mark Michelson","path":"/putnopvut","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1328590?s=80&v=4"},"commit":{"message":"Prepare for 23.06.5.\n\nAcked-by: Numan Siddique \nSigned-off-by: Mark Michelson ","shortMessageHtmlLink":"Prepare for 23.06.5."}},{"before":"5233cea52d1878786bc55ff5eccc309ea2667964","after":"d70ff69318f4ef4553ee1da32345d5243525b698","ref":"refs/heads/branch-23.09","pushedAt":"2024-05-10T17:33:54.000Z","pushType":"push","commitsCount":2,"pusher":{"login":"putnopvut","name":"Mark Michelson","path":"/putnopvut","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1328590?s=80&v=4"},"commit":{"message":"Prepare for 23.09.5.\n\nAcked-by: Numan Siddique \nSigned-off-by: Mark Michelson ","shortMessageHtmlLink":"Prepare for 23.09.5."}},{"before":"8cde6c5a8c308a6b293c9510b845de81a5afb932","after":"9c8264e765548f7c2f2e15af217c84abc664b932","ref":"refs/heads/branch-24.03","pushedAt":"2024-05-10T17:33:53.000Z","pushType":"push","commitsCount":2,"pusher":{"login":"putnopvut","name":"Mark Michelson","path":"/putnopvut","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1328590?s=80&v=4"},"commit":{"message":"Prepare for 24.03.3.\n\nAcked-by: Numan Siddique \nSigned-off-by: Mark Michelson ","shortMessageHtmlLink":"Prepare for 24.03.3."}},{"before":"36c5e15969698c284fcafe1ac9780c7446e118f5","after":"70c319085e0af16ab11de6d05d561e2dea991ce2","ref":"refs/heads/branch-23.03","pushedAt":"2024-05-10T16:33:24.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"controller: Fix an issue wrt cleanup of stale patch port.\n\nIssue:\nUpon updating the network_name option of localnet port from one physical\nbridge to another, ovn-controller fails to cleanup the peer localnet\npatch port from the old bridge and ends up creating a duplicate peer\nlocalnet patch port which fails in the following ovsdb transaction:\n\n```\n\"Transaction causes multiple rows in \\\"Interface\\\" table to have\nidentical values\n(patch-localnet_a7d47490-8a90-4c4d-8266-2915ad07c185-to-br-int)\n```\n\nCurrent workflow:\n1. Keep a set of all existing localnet ports on br-int. Let us call this\n set: existing_ports.\n2. For each localnet port in SB:\n 2.1 Create a patch port on br-int. (if it already exists on br-int,\n do not create but remove the entry from exisitng_ports set).\n 2.2 Create a peer patch port on br-phys-x. (if it already exists on the\n bridge, do not create but remove the entry from exisitng_ports set).\n3. Finally, cleanup all the ports and its peers from exisiting_ports.\n\nWith the above workflow, upon network_name change of localnet LSP, since\nports on br-int does not change, only peer port needs to be move from\nbr-phys-x to br-phys-y, the localnet port is removed from\nexisiting_ports in #2.1 and its peer never becomes eligible for cleanup.\n\nFix:\nInclude both patch port on br-int and its peer port in the\nexisiting_ports set as part of #1.\nThis make sures that the peer port is only removed from existing_ports\nonly if it is already present on the correct bridge. (#2.1/#2.2)\nOtherwise, during the cleanup in #3 it will be considered.\n\nFixes: b600316f252a (\"Don't delete patch ports that don't belong to br-int\")\nSigned-off-by: Priyankar Jain \nAcked-by: Numan Siddique \nSigned-off-by: Numan Siddique \n(cherry picked from commit 2609cd9f0b6e215780b08c8ec0a12fd0f043d5ca)","shortMessageHtmlLink":"controller: Fix an issue wrt cleanup of stale patch port."}},{"before":"88f8d8f2cbbd472fbdae0d89703d5578af4fad33","after":"920339ec76a7800455feeb731efa18b6f5ed896b","ref":"refs/heads/branch-23.06","pushedAt":"2024-05-10T16:33:09.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"controller: Fix an issue wrt cleanup of stale patch port.\n\nIssue:\nUpon updating the network_name option of localnet port from one physical\nbridge to another, ovn-controller fails to cleanup the peer localnet\npatch port from the old bridge and ends up creating a duplicate peer\nlocalnet patch port which fails in the following ovsdb transaction:\n\n```\n\"Transaction causes multiple rows in \\\"Interface\\\" table to have\nidentical values\n(patch-localnet_a7d47490-8a90-4c4d-8266-2915ad07c185-to-br-int)\n```\n\nCurrent workflow:\n1. Keep a set of all existing localnet ports on br-int. Let us call this\n set: existing_ports.\n2. For each localnet port in SB:\n 2.1 Create a patch port on br-int. (if it already exists on br-int,\n do not create but remove the entry from exisitng_ports set).\n 2.2 Create a peer patch port on br-phys-x. (if it already exists on the\n bridge, do not create but remove the entry from exisitng_ports set).\n3. Finally, cleanup all the ports and its peers from exisiting_ports.\n\nWith the above workflow, upon network_name change of localnet LSP, since\nports on br-int does not change, only peer port needs to be move from\nbr-phys-x to br-phys-y, the localnet port is removed from\nexisiting_ports in #2.1 and its peer never becomes eligible for cleanup.\n\nFix:\nInclude both patch port on br-int and its peer port in the\nexisiting_ports set as part of #1.\nThis make sures that the peer port is only removed from existing_ports\nonly if it is already present on the correct bridge. (#2.1/#2.2)\nOtherwise, during the cleanup in #3 it will be considered.\n\nFixes: b600316f252a (\"Don't delete patch ports that don't belong to br-int\")\nSigned-off-by: Priyankar Jain \nAcked-by: Numan Siddique \nSigned-off-by: Numan Siddique \n(cherry picked from commit 2609cd9f0b6e215780b08c8ec0a12fd0f043d5ca)","shortMessageHtmlLink":"controller: Fix an issue wrt cleanup of stale patch port."}},{"before":"b2d306be940456ffb15e014d49ae66516b16dcb1","after":"5233cea52d1878786bc55ff5eccc309ea2667964","ref":"refs/heads/branch-23.09","pushedAt":"2024-05-10T16:32:09.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"controller: Fix an issue wrt cleanup of stale patch port.\n\nIssue:\nUpon updating the network_name option of localnet port from one physical\nbridge to another, ovn-controller fails to cleanup the peer localnet\npatch port from the old bridge and ends up creating a duplicate peer\nlocalnet patch port which fails in the following ovsdb transaction:\n\n```\n\"Transaction causes multiple rows in \\\"Interface\\\" table to have\nidentical values\n(patch-localnet_a7d47490-8a90-4c4d-8266-2915ad07c185-to-br-int)\n```\n\nCurrent workflow:\n1. Keep a set of all existing localnet ports on br-int. Let us call this\n set: existing_ports.\n2. For each localnet port in SB:\n 2.1 Create a patch port on br-int. (if it already exists on br-int,\n do not create but remove the entry from exisitng_ports set).\n 2.2 Create a peer patch port on br-phys-x. (if it already exists on the\n bridge, do not create but remove the entry from exisitng_ports set).\n3. Finally, cleanup all the ports and its peers from exisiting_ports.\n\nWith the above workflow, upon network_name change of localnet LSP, since\nports on br-int does not change, only peer port needs to be move from\nbr-phys-x to br-phys-y, the localnet port is removed from\nexisiting_ports in #2.1 and its peer never becomes eligible for cleanup.\n\nFix:\nInclude both patch port on br-int and its peer port in the\nexisiting_ports set as part of #1.\nThis make sures that the peer port is only removed from existing_ports\nonly if it is already present on the correct bridge. (#2.1/#2.2)\nOtherwise, during the cleanup in #3 it will be considered.\n\nFixes: b600316f252a (\"Don't delete patch ports that don't belong to br-int\")\nSigned-off-by: Priyankar Jain \nAcked-by: Numan Siddique \nSigned-off-by: Numan Siddique \n(cherry picked from commit 2609cd9f0b6e215780b08c8ec0a12fd0f043d5ca)","shortMessageHtmlLink":"controller: Fix an issue wrt cleanup of stale patch port."}},{"before":"3a817c8684c30de0de35742753adb74c3a9b2f0f","after":"b2d306be940456ffb15e014d49ae66516b16dcb1","ref":"refs/heads/branch-23.09","pushedAt":"2024-05-10T16:27:43.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"utilities: Add ovn-debug binary tool.\n\nAdd ovn-debug binary tool that can be extended with commands that\nmight be useful for tests/debugging of OVN environment. Currently\nthe tool supports only two commands:\n\n1) \"lflow-stage-to-ltable STAGE_NAME\" that converts stage name into\n logical flow table id.\n\n2) \"lflow-stage-to-oftable STAGE_NAME\" that converts stage name into\n OpenFlow table id.\n\nFor now it will be used in tests to get rid remaining hardcoded table\nnumbers.\n\nNote: Will backporting this patch to branch-23.09, had to move the\novn stage macros from northd/northd.c to northd/northd.h in order to\nresolve the compilation errors in utilities/ovn-debug.c.\n\nSigned-off-by: Ales Musil \nAcked-by: Mark Michelson \nSigned-off-by: Mark Michelson \n(cherry picked from commit 117573e2d2790231c8ad5fbaba99b2a711c89e77)\nSigned-off-by: Numan Siddique ","shortMessageHtmlLink":"utilities: Add ovn-debug binary tool."}},{"before":"765f80c6470bf9e92f58ed57691cae303a2e6cd5","after":"8cde6c5a8c308a6b293c9510b845de81a5afb932","ref":"refs/heads/branch-24.03","pushedAt":"2024-05-10T16:19:00.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"controller: Fix an issue wrt cleanup of stale patch port.\n\nIssue:\nUpon updating the network_name option of localnet port from one physical\nbridge to another, ovn-controller fails to cleanup the peer localnet\npatch port from the old bridge and ends up creating a duplicate peer\nlocalnet patch port which fails in the following ovsdb transaction:\n\n```\n\"Transaction causes multiple rows in \\\"Interface\\\" table to have\nidentical values\n(patch-localnet_a7d47490-8a90-4c4d-8266-2915ad07c185-to-br-int)\n```\n\nCurrent workflow:\n1. Keep a set of all existing localnet ports on br-int. Let us call this\n set: existing_ports.\n2. For each localnet port in SB:\n 2.1 Create a patch port on br-int. (if it already exists on br-int,\n do not create but remove the entry from exisitng_ports set).\n 2.2 Create a peer patch port on br-phys-x. (if it already exists on the\n bridge, do not create but remove the entry from exisitng_ports set).\n3. Finally, cleanup all the ports and its peers from exisiting_ports.\n\nWith the above workflow, upon network_name change of localnet LSP, since\nports on br-int does not change, only peer port needs to be move from\nbr-phys-x to br-phys-y, the localnet port is removed from\nexisiting_ports in #2.1 and its peer never becomes eligible for cleanup.\n\nFix:\nInclude both patch port on br-int and its peer port in the\nexisiting_ports set as part of #1.\nThis make sures that the peer port is only removed from existing_ports\nonly if it is already present on the correct bridge. (#2.1/#2.2)\nOtherwise, during the cleanup in #3 it will be considered.\n\nFixes: b600316f252a (\"Don't delete patch ports that don't belong to br-int\")\nSigned-off-by: Priyankar Jain \nAcked-by: Numan Siddique \nSigned-off-by: Numan Siddique \n(cherry picked from commit 2609cd9f0b6e215780b08c8ec0a12fd0f043d5ca)","shortMessageHtmlLink":"controller: Fix an issue wrt cleanup of stale patch port."}},{"before":"40a268313c35751d65c35e6d13154a99a3c6069f","after":"765f80c6470bf9e92f58ed57691cae303a2e6cd5","ref":"refs/heads/branch-24.03","pushedAt":"2024-05-10T15:55:49.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"utilities: Add ovn-debug binary tool.\n\nAdd ovn-debug binary tool that can be extended with commands that\nmight be useful for tests/debugging of OVN environment. Currently\nthe tool supports only two commands:\n\n1) \"lflow-stage-to-ltable STAGE_NAME\" that converts stage name into\n logical flow table id.\n\n2) \"lflow-stage-to-oftable STAGE_NAME\" that converts stage name into\n OpenFlow table id.\n\nFor now it will be used in tests to get rid remaining hardcoded table\nnumbers.\n\nSigned-off-by: Ales Musil \nAcked-by: Mark Michelson \nSigned-off-by: Mark Michelson \n(cherry picked from commit 117573e2d2790231c8ad5fbaba99b2a711c89e77)","shortMessageHtmlLink":"utilities: Add ovn-debug binary tool."}},{"before":"465d2db2f3e518907879fe400e6d429aa095ad74","after":"3a817c8684c30de0de35742753adb74c3a9b2f0f","ref":"refs/heads/branch-23.09","pushedAt":"2024-05-10T06:37:07.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"hzhou8","name":"Han Zhou","path":"/hzhou8","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7600054?s=80&v=4"},"commit":{"message":"controller: Track individual address set constants.\n\nInstead of tracking address set per struct expr_constant_set track it\nper individual struct expr_constant. This allows more fine grained\ncontrol for I-P processing of address sets in controller. It helps with\nscenarios like matching on two address sets in one expression e.g.\n\"ip4.src == {$as1, $as2}\". This allows any addition or removal of\nindividual adress from the set to be incrementally processed instead\nof reprocessing all the flows.\n\nThis unfortunately doesn't help with the following case:\n\"ip4.src == $as1 || ip4.dst == $as2\"\n\nThe memory impact should be minimal as there is only increase of 8 bytes\nper the struct expr_constant.\n\nReported-at: https://issues.redhat.com/browse/FDP-509\nSigned-off-by: Ales Musil \nSigned-off-by: Han Zhou \n(cherry picked from commit cce476ce39e7eb3d3676c40e2d54af8668dd7868)\n(cherry picked from commit 40a268313c35751d65c35e6d13154a99a3c6069f)","shortMessageHtmlLink":"controller: Track individual address set constants."}},{"before":"20ab6c83f57fe14154f5b117138c512fc6b49388","after":"40a268313c35751d65c35e6d13154a99a3c6069f","ref":"refs/heads/branch-24.03","pushedAt":"2024-05-10T06:06:04.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"hzhou8","name":"Han Zhou","path":"/hzhou8","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7600054?s=80&v=4"},"commit":{"message":"controller: Track individual address set constants.\n\nInstead of tracking address set per struct expr_constant_set track it\nper individual struct expr_constant. This allows more fine grained\ncontrol for I-P processing of address sets in controller. It helps with\nscenarios like matching on two address sets in one expression e.g.\n\"ip4.src == {$as1, $as2}\". This allows any addition or removal of\nindividual adress from the set to be incrementally processed instead\nof reprocessing all the flows.\n\nThis unfortunately doesn't help with the following case:\n\"ip4.src == $as1 || ip4.dst == $as2\"\n\nThe memory impact should be minimal as there is only increase of 8 bytes\nper the struct expr_constant.\n\nReported-at: https://issues.redhat.com/browse/FDP-509\nSigned-off-by: Ales Musil \nSigned-off-by: Han Zhou \n(cherry picked from commit cce476ce39e7eb3d3676c40e2d54af8668dd7868)","shortMessageHtmlLink":"controller: Track individual address set constants."}},{"before":"0cfe12a7901d6f5189675d5fe76a021a798124a1","after":"20ab6c83f57fe14154f5b117138c512fc6b49388","ref":"refs/heads/branch-24.03","pushedAt":"2024-05-09T23:37:56.000Z","pushType":"push","commitsCount":2,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"northd, controller: Handle tunnel_key change consistently.\n\nCurrently the tunnel_key change for either LS/LR/LSP/LRP wasn't\nconsistent. That would lead to a situations when some old would still\nbe present, breaking the connection especially for already existing\nFDBs and MAC bindings.\n\nMake sure the FDB entries are up to date by removing them from DB\nwhen there is a tunnel_key change as those entries have only tunnel_key\nrefrences (dp_key, port_key).\n\nMAC bindings have references to the datapath and port name, instead of\nremoving those entries do recompute in the controller when we detect\ntunnel_key change. This can be costly at scale, however the tunnel_key\nis not expected to change constantly, in most cases it shouldn't change\nat all.\n\nFixes: b337750e45be (\"northd: Incremental processing of VIF changes in 'northd' node.\")\nFixes: 425f699e2b20 (\"controller: fixed potential segfault when changing tunnel_key and deleting ls.\")\nReported-at: https://issues.redhat.com/browse/FDP-393\nAcked-by: Mark Michelson \nSigned-off-by: Ales Musil \nSigned-off-by: Numan Siddique \n(cherry picked from commit ddf051cbc6af24c303bf88970750e5c5fe285400)","shortMessageHtmlLink":"northd, controller: Handle tunnel_key change consistently."}},{"before":"b994d1d3ba071e749ff0a9441fc15305d7384a9d","after":"2609cd9f0b6e215780b08c8ec0a12fd0f043d5ca","ref":"refs/heads/main","pushedAt":"2024-05-09T23:33:44.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"numansiddique","name":null,"path":"/numansiddique","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/6571859?s=80&v=4"},"commit":{"message":"controller: Fix an issue wrt cleanup of stale patch port.\n\nIssue:\nUpon updating the network_name option of localnet port from one physical\nbridge to another, ovn-controller fails to cleanup the peer localnet\npatch port from the old bridge and ends up creating a duplicate peer\nlocalnet patch port which fails in the following ovsdb transaction:\n\n```\n\"Transaction causes multiple rows in \\\"Interface\\\" table to have\nidentical values\n(patch-localnet_a7d47490-8a90-4c4d-8266-2915ad07c185-to-br-int)\n```\n\nCurrent workflow:\n1. Keep a set of all existing localnet ports on br-int. Let us call this\n set: existing_ports.\n2. For each localnet port in SB:\n 2.1 Create a patch port on br-int. (if it already exists on br-int,\n do not create but remove the entry from exisitng_ports set).\n 2.2 Create a peer patch port on br-phys-x. (if it already exists on the\n bridge, do not create but remove the entry from exisitng_ports set).\n3. Finally, cleanup all the ports and its peers from exisiting_ports.\n\nWith the above workflow, upon network_name change of localnet LSP, since\nports on br-int does not change, only peer port needs to be move from\nbr-phys-x to br-phys-y, the localnet port is removed from\nexisiting_ports in #2.1 and its peer never becomes eligible for cleanup.\n\nFix:\nInclude both patch port on br-int and its peer port in the\nexisiting_ports set as part of #1.\nThis make sures that the peer port is only removed from existing_ports\nonly if it is already present on the correct bridge. (#2.1/#2.2)\nOtherwise, during the cleanup in #3 it will be considered.\n\nFixes: b600316f252a (\"Don't delete patch ports that don't belong to br-int\")\nSigned-off-by: Priyankar Jain \nAcked-by: Numan Siddique \nSigned-off-by: Numan Siddique ","shortMessageHtmlLink":"controller: Fix an issue wrt cleanup of stale patch port."}}],"hasNextPage":true,"hasPreviousPage":false,"activityType":"all","actor":null,"timePeriod":"all","sort":"DESC","perPage":30,"cursor":"djE6ks8AAAAEUEUd1QA","startCursor":null,"endCursor":null}},"title":"Activity ยท ovn-org/ovn"}