Koordinator Files

What's Changed

• scheduler: refine reservation msg if matched filtered out by other plugins by @saintube in https://github.com/koordinator-sh/koordinator/pull/2085
• Add ZiMengSheng into approvers alias by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2087
• scheduler: fix reservation plugin clone concurrent read write by @buptcozy in https://github.com/koordinator-sh/koordinator/pull/2084
• scheduler: fix elastic quota TestPlugin_OnPodDelete random panic by @buptcozy in https://github.com/koordinator-sh/koordinator/pull/2083
• scheduler: change reservation event message format by @zwzhang0107 in https://github.com/koordinator-sh/koordinator/pull/2090
• scheduler: add node info in reservation level event by @zwzhang0107 in https://github.com/koordinator-sh/koordinator/pull/2094
• scheduler: support consume reserved numa resource by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2080
• scheduler: add new pod estimate with loadaware plugin by @zwForrest in https://github.com/koordinator-sh/koordinator/pull/1992
• koord-descheduler: limit the total number of pod evictions by @zwForrest in https://github.com/koordinator-sh/koordinator/pull/2091
• chore: fix e2e configs after the load-aware scheduling updated by @saintube in https://github.com/koordinator-sh/koordinator/pull/2101
• koord-descheduler: balance prod pods between nodes in LowNodeLoad by @zwForrest in https://github.com/koordinator-sh/koordinator/pull/2066
• chore: fix low node load test after prod thresholds added by @songtao98 in https://github.com/koordinator-sh/koordinator/pull/2103
• koord-scheduler: Pod updates should not update the timestamp by @zwForrest in https://github.com/koordinator-sh/koordinator/pull/2100
• koord-descheduler: fix object limiter by move it to reconciler by @songtao98 in https://github.com/koordinator-sh/koordinator/pull/2088
• koord-descheduler: enhance LowNodeLoad scorer by @LY-today in https://github.com/koordinator-sh/koordinator/pull/2092
• koord-descheduler: support evicting all bare pods in migration controller by @songtao98 in https://github.com/koordinator-sh/koordinator/pull/2102
• koordlet: update resctrl qos proposal by @kangclzjc in https://github.com/koordinator-sh/koordinator/pull/2079
• koordlet: add taskids in statesinformer by @kangclzjc in https://github.com/koordinator-sh/koordinator/pull/2057
• koord-manager: fix the prom metrics handler by @saintube in https://github.com/koordinator-sh/koordinator/pull/2107
• koord-descheduler: fix descheduler log by @zwForrest in https://github.com/koordinator-sh/koordinator/pull/2114
• scheduler: support pod request exact match reservation by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2121
• scheduler: reduce cycleState overhead for reservation by @saintube in https://github.com/koordinator-sh/koordinator/pull/2120
• scheduler: fix reserveNUMAResource bug by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2124
• koordlet: add allpods reconciler func by @kangclzjc in https://github.com/koordinator-sh/koordinator/pull/2112
• scheduler: reduce deviceShare memory overhead by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2126
• webhook: add quota evaluate webhook by @shaloulcy in https://github.com/koordinator-sh/koordinator/pull/2129
• koordlet: upgrade nri to 0.6.1 by @kangclzjc in https://github.com/koordinator-sh/koordinator/pull/2132
• koord-manager: support manually configured default cpu normalization settings by @yangfeiyu20102011 in https://github.com/koordinator-sh/koordinator/pull/2128
• koordlet: tc plugin for netqos by @lucming in https://github.com/koordinator-sh/koordinator/pull/1920
• chore: fix koordlet dockerfile by @saintube in https://github.com/koordinator-sh/koordinator/pull/2134
• koord-descheduler: Fix threshold initialization for prod. by @zwForrest in https://github.com/koordinator-sh/koordinator/pull/2130
• scheduler: support device topology strategy by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2133
• scheduler: optimize when only 1 reservation/node by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2136
• koord-descheduler: add migration object limiter for namespace by @songtao98 in https://github.com/koordinator-sh/koordinator/pull/2068
• webhook: revise elasticquota mutating when validating disabled by @saintube in https://github.com/koordinator-sh/koordinator/pull/2135
• koord-descheduler: support max migrating globally limitation by @songtao98 in https://github.com/koordinator-sh/koordinator/pull/2143
• scheduler: skip uncessary Filter&Score plugin by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2138
• all: support priority and preemption policy transformer by @saintube in https://github.com/koordinator-sh/koordinator/pull/2137
• scheduler: elastic quota ignore terminating pod by @shaloulcy in https://github.com/koordinator-sh/koordinator/pull/2141
• koord-descheduler: fix maxMigratingGlobally arg json tag by @songtao98 in https://github.com/koordinator-sh/koordinator/pull/2144
• scheduler:support multi gpu share by @AdrianMachao in https://github.com/koordinator-sh/koordinator/pull/2127
• scheduler: return framework.Unschedulable when cpuPolicy not satisfied by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2146
• koord-manager: fix invalid assignments in the resource amplification … by @yangfeiyu20102011 in https://github.com/koordinator-sh/koordinator/pull/2149
• koordlet: add some information to improve log readability by @yangfeiyu20102011 in https://github.com/koordinator-sh/koordinator/pull/2153
• ci: add disk GC for E2E jobs by @saintube in https://github.com/koordinator-sh/koordinator/pull/2157
• scheduler: fix panic when NUMANode equals -1 by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2151
• scheduler: fix too many ut log by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2154
• scheduler: support scheduler config v1 by @AdrianMachao in https://github.com/koordinator-sh/koordinator/pull/2167
• scheduler: add coscheduling preEnqueue by @AdrianMachao in https://github.com/koordinator-sh/koordinator/pull/2155
• koordlet: add record events by @kangclzjc in https://github.com/koordinator-sh/koordinator/pull/2162
• koordlet: fix unsafe conversion in net_cls by @saintube in https://github.com/koordinator-sh/koordinator/pull/2173
• scheduler: assure only gpu will allocate by topology by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2178
• schedler: move gang OnceSatified to gangGroupInfo by @buptcozy in https://github.com/koordinator-sh/koordinator/pull/2176
• scheduler: add reservation preemption by @saintube in https://github.com/koordinator-sh/koordinator/pull/2139
• koord-descheduler: update k8s descheduler to 0.28.0 by @songtao98 in https://github.com/koordinator-sh/koordinator/pull/2156
• scheduler: support reservation ignored and nodenumaresource preemption by @saintube in https://github.com/koordinator-sh/koordinator/pull/2163
• koord-manager: support resource amplification config, cpu, memory and other resource by @yangfeiyu20102011 in https://github.com/koordinator-sh/koordinator/pull/2172
• scheduler: fix pod update when pod creation is before quota creation by @shaloulcy in https://github.com/koordinator-sh/koordinator/pull/2177
• scheduler: permit other plugin to set numaAffinity by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2182
• koord-descheduler: LowNodeLoad check if evicted pod can cause new node over utilized by @songtao98 in https://github.com/koordinator-sh/koordinator/pull/2142
• Update OWNERS_ALIASES by @hormes in https://github.com/koordinator-sh/koordinator/pull/2145
• koord-descheduler: fixes namespace object limiter by @songtao98 in https://github.com/koordinator-sh/koordinator/pull/2160
• scheduler: suppport amd.com/gpu by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2174
• scheduler: elastic quota ignore terminating pod immediately by @TaoYang526 in https://github.com/koordinator-sh/koordinator/pull/2180
• apis: add reservation name, taints and tolerations in ReservationAffinitty by @zwzhang0107 in https://github.com/koordinator-sh/koordinator/pull/2186
• scheduler: fix panic when nonimatingInfo nil by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2189
• apis: add protobuf for reservation by @zwzhang0107 in https://github.com/koordinator-sh/koordinator/pull/2192
• scheduler: fix ElasticQuota state's Clone by @saintube in https://github.com/koordinator-sh/koordinator/pull/2190
• koordlet: fix wrong msg in calculateBESuppressCPU by @yangfeiyu20102011 in https://github.com/koordinator-sh/koordinator/pull/2193
• chores(deps): Add support depguard rules in golangci-lint by @dongjiang1989 in https://github.com/koordinator-sh/koordinator/pull/1965
• koord-descheduler: fix migration controller max unavailable computing algorigthm by @songtao98 in https://github.com/koordinator-sh/koordinator/pull/2196
• scheduler: support devices of the same node gpuMem not equal by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2199
• scheduler: support pod preemption from numa awareless reservation by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2204
• scheduler: optimize numa affinity store by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2209
• apis: fix type of MinResources in podgroup by @zwzhang0107 in https://github.com/koordinator-sh/koordinator/pull/2212
• koordlet: Add resctrl runtime hook for pod level by @kangclzjc in https://github.com/koordinator-sh/koordinator/pull/2123
• koord-manager: nodeslo-controller enqueue request when node labels updated by @chengjoey in https://github.com/koordinator-sh/koordinator/pull/2201
• scheduler: optimize performance on transformer extension and Skip status by @saintube in https://github.com/koordinator-sh/koordinator/pull/2211
• koord-manager: add unallocated resource into mid resource. by @tan90github in https://github.com/koordinator-sh/koordinator/pull/2152
• scheduler: support reservation name, taints and tolerations by @saintube in https://github.com/koordinator-sh/koordinator/pull/2207
• scheduler: recover gang check in preFilter by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2217
• apis: GPU Partition related by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2219
• scheduler: optimize GPU allocate logic by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2221
• scheduler: support amd by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2223
• scheduler: support reserving pods resources by @saintube in https://github.com/koordinator-sh/koordinator/pull/2224
• scheduler: allocate tolerate numa-meaning-less device by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2226
• scheduler: fix partition binpack disorder by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2227
• koord-descheduler: fix prod pod excessive eviction when node recover to normal by @JBinin in https://github.com/koordinator-sh/koordinator/pull/2225
• scheduler: fix gpu shared bug by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2228
• scheduler: support secondary device well planned by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2229
• scheduler: fix shared gpu pod allocated minor of -1 by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2230
• scheduler: improve gang log by @googs1025 in https://github.com/koordinator-sh/koordinator/pull/2009
• scheduler: consider pod requests when gpu&RDMA joint allocate by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2233
• scheduler: optimize reservation perf with lazy restoring by @saintube in https://github.com/koordinator-sh/koordinator/pull/2241
• util: fix incorrect comparison for resource.go#LessThanOrEqualCompletely by @TaoYang526 in https://github.com/koordinator-sh/koordinator/pull/2235
• scheduler: support besteffort policy by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2243
• scheduler: fix scaling factor 100 for burstable pod by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2242
• gpu: setting default gpu partition policy by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2245
• koordlet: add ReadMemoryUsage method to improve generality of cgroup reader interface by @yangfeiyu20102011 in https://github.com/koordinator-sh/koordinator/pull/2251
• koord-manager: refactor oversale resource calculate logic. by @tan90github in https://github.com/koordinator-sh/koordinator/pull/2240
• util: add defaulting to blkio qos to improve robustness by @zqzten in https://github.com/koordinator-sh/koordinator/pull/2238
• webhook: Support setting quota admission to zero for special usage scenarios, such as temporarily pausing pod submissions. by @TaoYang526 in https://github.com/koordinator-sh/koordinator/pull/2237
• scheduler: add downgrade strategy for empty 'aggregated' on cold koor… by @clay-wangzhi in https://github.com/koordinator-sh/koordinator/pull/2239
• koord-manager: mv slo-controller metrics to util by @zwzhang0107 in https://github.com/koordinator-sh/koordinator/pull/2257
• scheduler: fix deviceshare with reservation-ignored pods by @saintube in https://github.com/koordinator-sh/koordinator/pull/2252
• scheduler: setFailedPlugin when plugin transformer failed by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2262
• koordlet: export host application cpu and memory usage for prometheus by @yangfeiyu20102011 in https://github.com/koordinator-sh/koordinator/pull/2259
• scheduler: update dingtalk QR Code by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2277
• scheduler: support allocating from reservation when no resource matched by @saintube in https://github.com/koordinator-sh/koordinator/pull/2279
• koordlet: add GetNodeMetricSpec in statesInformer interface by @j4ckstraw in https://github.com/koordinator-sh/koordinator/pull/2278
• api: supply id for minor meaningless device by @ferris-cx in https://github.com/koordinator-sh/koordinator/pull/2250
• koordlet: remove useless stateInformer.impl StateInformer interface by @j4ckstraw in https://github.com/koordinator-sh/koordinator/pull/2286
• Upgrade ginkgo to v2.11.0 by @nce3xin in https://github.com/koordinator-sh/koordinator/pull/2284
• koordlet: support change pod CPUQOS by annotations by @j4ckstraw in https://github.com/koordinator-sh/koordinator/pull/2206
• koordlet: rdma device inject by @ferris-cx in https://github.com/koordinator-sh/koordinator/pull/2285
• koord-manager: consider NodeReserved when calculate mid resource. by @tan90github in https://github.com/koordinator-sh/koordinator/pull/2253
• koordlet: fix reconciler description mismatch. by @tan90github in https://github.com/koordinator-sh/koordinator/pull/2282
• manager: sync rdma resource to node by @ferris-cx in https://github.com/koordinator-sh/koordinator/pull/2249
• koordlet: supply rdma devices by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2276
• koordlet: add metrics about be used cpu and node used memory by @leason00 in https://github.com/koordinator-sh/koordinator/pull/2283
• scheduler: add log for cpu/mem numaaware allocate by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2293
• koordlet: fix GeviceNumbers compatibility in mac by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2295
• koordlet: collect metric for host application memory usage with page cache by @j4ckstraw in https://github.com/koordinator-sh/koordinator/pull/2273
• slo-controller: fix mid resource calculate formula by @lijunxin559 in https://github.com/koordinator-sh/koordinator/pull/2291
• gpu: support strict gpu share with hami by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2272
• scheduler: make error more clear when numa-aware by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2305
• scheduler: fix singleNUMANodeExclusive not clone by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2309
• scheduler: fix UT Fail cause map iterate by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2311
• scheduler: refactor reset quota by @shaloulcy in https://github.com/koordinator-sh/koordinator/pull/2301
• scheduler: extend schedulerMonitor by @saintube in https://github.com/koordinator-sh/koordinator/pull/2314
• scheduler: register pod delete event for resource plugin by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2308
• scheduler: delete quota inplace by @shaloulcy in https://github.com/koordinator-sh/koordinator/pull/2312
• scheduler: add quota inplace by @shaloulcy in https://github.com/koordinator-sh/koordinator/pull/2313
• scheduler: revise ReservationFilterPlugin and fix preempting pods resources by @saintube in https://github.com/koordinator-sh/koordinator/pull/2315
• scheduler: fix ReplaceQuotas ut by @shaloulcy in https://github.com/koordinator-sh/koordinator/pull/2318
• scheduler: make topology manager aware device preference by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2316
• scheduler: fix quota check panic by @shaloulcy in https://github.com/koordinator-sh/koordinator/pull/2321
• scheduler: fix reservation reserved updated by @saintube in https://github.com/koordinator-sh/koordinator/pull/2322
• scheduler: export GPUPartitionIndexOfNVIDIAHopper by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2324
• koordlet: fix prodReclaimablePredictor result to avoid influence of o… by @lijunxin559 in https://github.com/koordinator-sh/koordinator/pull/2325
• feat: add NodeResourcesFitPlus and ScarceResourceAvoidance plugin by @LY-today in https://github.com/koordinator-sh/koordinator/pull/2302
• koord-manager: add metrics for webhook by @nce3xin in https://github.com/koordinator-sh/koordinator/pull/2330
• scheduler: fix monitor terminating pod by @saintube in https://github.com/koordinator-sh/koordinator/pull/2331
• koordlet: pod resources proxy by @ferris-cx in https://github.com/koordinator-sh/koordinator/pull/2300
• scheduler: some minor optimizations for reservation by @saintube in https://github.com/koordinator-sh/koordinator/pull/2336
• scheduler: move custom informers for extension by @saintube in https://github.com/koordinator-sh/koordinator/pull/2338
• webhook: support elasticquota enable update resource key by @lijunxin559 in https://github.com/koordinator-sh/koordinator/pull/2323
• docs: pod resources proxy by @ferris-cx in https://github.com/koordinator-sh/koordinator/pull/2292
• koordlet: fix podresources not found by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2344
• koordlet: fix UT for podresources by @ZiMengSheng in https://github.com/koordinator-sh/koordinator/pull/2345

New Contributors

• @LY-today made their first contribution in https://github.com/koordinator-sh/koordinator/pull/2092
• @AdrianMachao made their first contribution in https://github.com/koordinator-sh/koordinator/pull/2127
• @TaoYang526 made their first contribution in https://github.com/koordinator-sh/koordinator/pull/2180
• @dongjiang1989 made their first contribution in https://github.com/koordinator-sh/koordinator/pull/1965
• @chengjoey made their first contribution in https://github.com/koordinator-sh/koordinator/pull/2201
• @JBinin made their first contribution in https://github.com/koordinator-sh/koordinator/pull/2225
• @clay-wangzhi made their first contribution in https://github.com/koordinator-sh/koordinator/pull/2239
• @ferris-cx made their first contribution in https://github.com/koordinator-sh/koordinator/pull/2250
• @nce3xin made their first contribution in https://github.com/koordinator-sh/koordinator/pull/2284
• @lijunxin559 made their first contribution in https://github.com/koordinator-sh/koordinator/pull/2291

Full Changelog: https://github.com/koordinator-sh/koordinator/compare/v1.5.0...v1.6.0