skypilot-nightly 1.0.0.dev20250509__py3-none-any.whl → 1.0.0.dev20251107__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.

Potentially problematic release.


This version of skypilot-nightly might be problematic. Click here for more details.

Files changed (512) hide show
  1. sky/__init__.py +22 -6
  2. sky/adaptors/aws.py +25 -7
  3. sky/adaptors/common.py +24 -1
  4. sky/adaptors/coreweave.py +278 -0
  5. sky/adaptors/do.py +8 -2
  6. sky/adaptors/hyperbolic.py +8 -0
  7. sky/adaptors/kubernetes.py +149 -18
  8. sky/adaptors/nebius.py +170 -17
  9. sky/adaptors/primeintellect.py +1 -0
  10. sky/adaptors/runpod.py +68 -0
  11. sky/adaptors/seeweb.py +167 -0
  12. sky/adaptors/shadeform.py +89 -0
  13. sky/admin_policy.py +187 -4
  14. sky/authentication.py +179 -225
  15. sky/backends/__init__.py +4 -2
  16. sky/backends/backend.py +22 -9
  17. sky/backends/backend_utils.py +1299 -380
  18. sky/backends/cloud_vm_ray_backend.py +1715 -518
  19. sky/backends/docker_utils.py +1 -1
  20. sky/backends/local_docker_backend.py +11 -6
  21. sky/backends/wheel_utils.py +37 -9
  22. sky/{clouds/service_catalog → catalog}/__init__.py +21 -19
  23. sky/{clouds/service_catalog → catalog}/aws_catalog.py +27 -8
  24. sky/{clouds/service_catalog → catalog}/azure_catalog.py +10 -7
  25. sky/{clouds/service_catalog → catalog}/common.py +89 -48
  26. sky/{clouds/service_catalog → catalog}/cudo_catalog.py +8 -5
  27. sky/{clouds/service_catalog → catalog}/data_fetchers/analyze.py +1 -1
  28. sky/{clouds/service_catalog → catalog}/data_fetchers/fetch_aws.py +30 -40
  29. sky/{clouds/service_catalog → catalog}/data_fetchers/fetch_cudo.py +38 -38
  30. sky/{clouds/service_catalog → catalog}/data_fetchers/fetch_gcp.py +42 -15
  31. sky/catalog/data_fetchers/fetch_hyperbolic.py +136 -0
  32. sky/{clouds/service_catalog → catalog}/data_fetchers/fetch_lambda_cloud.py +1 -0
  33. sky/catalog/data_fetchers/fetch_nebius.py +335 -0
  34. sky/catalog/data_fetchers/fetch_runpod.py +698 -0
  35. sky/catalog/data_fetchers/fetch_seeweb.py +329 -0
  36. sky/catalog/data_fetchers/fetch_shadeform.py +142 -0
  37. sky/{clouds/service_catalog → catalog}/data_fetchers/fetch_vast.py +1 -1
  38. sky/{clouds/service_catalog → catalog}/data_fetchers/fetch_vsphere.py +1 -1
  39. sky/{clouds/service_catalog → catalog}/do_catalog.py +5 -2
  40. sky/{clouds/service_catalog → catalog}/fluidstack_catalog.py +6 -3
  41. sky/{clouds/service_catalog → catalog}/gcp_catalog.py +41 -15
  42. sky/catalog/hyperbolic_catalog.py +136 -0
  43. sky/{clouds/service_catalog → catalog}/ibm_catalog.py +9 -6
  44. sky/{clouds/service_catalog → catalog}/kubernetes_catalog.py +36 -24
  45. sky/{clouds/service_catalog → catalog}/lambda_catalog.py +9 -6
  46. sky/{clouds/service_catalog → catalog}/nebius_catalog.py +9 -7
  47. sky/{clouds/service_catalog → catalog}/oci_catalog.py +9 -6
  48. sky/{clouds/service_catalog → catalog}/paperspace_catalog.py +5 -2
  49. sky/catalog/primeintellect_catalog.py +95 -0
  50. sky/{clouds/service_catalog → catalog}/runpod_catalog.py +11 -4
  51. sky/{clouds/service_catalog → catalog}/scp_catalog.py +9 -6
  52. sky/catalog/seeweb_catalog.py +184 -0
  53. sky/catalog/shadeform_catalog.py +165 -0
  54. sky/catalog/ssh_catalog.py +167 -0
  55. sky/{clouds/service_catalog → catalog}/vast_catalog.py +6 -3
  56. sky/{clouds/service_catalog → catalog}/vsphere_catalog.py +5 -2
  57. sky/check.py +491 -203
  58. sky/cli.py +5 -6005
  59. sky/client/{cli.py → cli/command.py} +2477 -1885
  60. sky/client/cli/deprecation_utils.py +99 -0
  61. sky/client/cli/flags.py +359 -0
  62. sky/client/cli/table_utils.py +320 -0
  63. sky/client/common.py +70 -32
  64. sky/client/oauth.py +82 -0
  65. sky/client/sdk.py +1203 -297
  66. sky/client/sdk_async.py +833 -0
  67. sky/client/service_account_auth.py +47 -0
  68. sky/cloud_stores.py +73 -0
  69. sky/clouds/__init__.py +13 -0
  70. sky/clouds/aws.py +358 -93
  71. sky/clouds/azure.py +105 -83
  72. sky/clouds/cloud.py +127 -36
  73. sky/clouds/cudo.py +68 -50
  74. sky/clouds/do.py +66 -48
  75. sky/clouds/fluidstack.py +63 -44
  76. sky/clouds/gcp.py +339 -110
  77. sky/clouds/hyperbolic.py +293 -0
  78. sky/clouds/ibm.py +70 -49
  79. sky/clouds/kubernetes.py +563 -162
  80. sky/clouds/lambda_cloud.py +74 -54
  81. sky/clouds/nebius.py +206 -80
  82. sky/clouds/oci.py +88 -66
  83. sky/clouds/paperspace.py +61 -44
  84. sky/clouds/primeintellect.py +317 -0
  85. sky/clouds/runpod.py +164 -74
  86. sky/clouds/scp.py +89 -83
  87. sky/clouds/seeweb.py +466 -0
  88. sky/clouds/shadeform.py +400 -0
  89. sky/clouds/ssh.py +263 -0
  90. sky/clouds/utils/aws_utils.py +10 -4
  91. sky/clouds/utils/gcp_utils.py +87 -11
  92. sky/clouds/utils/oci_utils.py +38 -14
  93. sky/clouds/utils/scp_utils.py +177 -124
  94. sky/clouds/vast.py +99 -77
  95. sky/clouds/vsphere.py +51 -40
  96. sky/core.py +349 -139
  97. sky/dag.py +15 -0
  98. sky/dashboard/out/404.html +1 -1
  99. sky/dashboard/out/_next/static/chunks/1141-e6aa9ab418717c59.js +11 -0
  100. sky/dashboard/out/_next/static/chunks/1272-1ef0bf0237faccdb.js +1 -0
  101. sky/dashboard/out/_next/static/chunks/1871-74503c8e80fd253b.js +6 -0
  102. sky/dashboard/out/_next/static/chunks/2260-7703229c33c5ebd5.js +1 -0
  103. sky/dashboard/out/_next/static/chunks/2350.fab69e61bac57b23.js +1 -0
  104. sky/dashboard/out/_next/static/chunks/2369.fc20f0c2c8ed9fe7.js +15 -0
  105. sky/dashboard/out/_next/static/chunks/2755.fff53c4a3fcae910.js +26 -0
  106. sky/dashboard/out/_next/static/chunks/3294.72362fa129305b19.js +1 -0
  107. sky/dashboard/out/_next/static/chunks/3785.ad6adaa2a0fa9768.js +1 -0
  108. sky/dashboard/out/_next/static/chunks/3850-ff4a9a69d978632b.js +1 -0
  109. sky/dashboard/out/_next/static/chunks/3937.210053269f121201.js +1 -0
  110. sky/dashboard/out/_next/static/chunks/4725.a830b5c9e7867c92.js +1 -0
  111. sky/dashboard/out/_next/static/chunks/4937.a2baa2df5572a276.js +15 -0
  112. sky/dashboard/out/_next/static/chunks/5739-d67458fcb1386c92.js +8 -0
  113. sky/dashboard/out/_next/static/chunks/6130-2be46d70a38f1e82.js +1 -0
  114. sky/dashboard/out/_next/static/chunks/616-3d59f75e2ccf9321.js +39 -0
  115. sky/dashboard/out/_next/static/chunks/6212-7bd06f60ba693125.js +13 -0
  116. sky/dashboard/out/_next/static/chunks/6601-06114c982db410b6.js +1 -0
  117. sky/dashboard/out/_next/static/chunks/6856-ef8ba11f96d8c4a3.js +1 -0
  118. sky/dashboard/out/_next/static/chunks/6989-01359c57e018caa4.js +1 -0
  119. sky/dashboard/out/_next/static/chunks/6990-32b6e2d3822301fa.js +1 -0
  120. sky/dashboard/out/_next/static/chunks/7359-c8d04e06886000b3.js +30 -0
  121. sky/dashboard/out/_next/static/chunks/7411-b15471acd2cba716.js +41 -0
  122. sky/dashboard/out/_next/static/chunks/7615-3301e838e5f25772.js +1 -0
  123. sky/dashboard/out/_next/static/chunks/8640.5b9475a2d18c5416.js +16 -0
  124. sky/dashboard/out/_next/static/chunks/8969-1e4613c651bf4051.js +1 -0
  125. sky/dashboard/out/_next/static/chunks/9025.fa408f3242e9028d.js +6 -0
  126. sky/dashboard/out/_next/static/chunks/9353-cff34f7e773b2e2b.js +1 -0
  127. sky/dashboard/out/_next/static/chunks/9360.7310982cf5a0dc79.js +31 -0
  128. sky/dashboard/out/_next/static/chunks/9847.3aaca6bb33455140.js +30 -0
  129. sky/dashboard/out/_next/static/chunks/fd9d1056-86323a29a8f7e46a.js +1 -0
  130. sky/dashboard/out/_next/static/chunks/framework-cf60a09ccd051a10.js +33 -0
  131. sky/dashboard/out/_next/static/chunks/main-app-587214043926b3cc.js +1 -0
  132. sky/dashboard/out/_next/static/chunks/main-f15ccb73239a3bf1.js +1 -0
  133. sky/dashboard/out/_next/static/chunks/pages/_app-bde01e4a2beec258.js +34 -0
  134. sky/dashboard/out/_next/static/chunks/pages/_error-c66a4e8afc46f17b.js +1 -0
  135. sky/dashboard/out/_next/static/chunks/pages/clusters/[cluster]/[job]-c736ead69c2d86ec.js +16 -0
  136. sky/dashboard/out/_next/static/chunks/pages/clusters/[cluster]-a37d2063af475a1c.js +1 -0
  137. sky/dashboard/out/_next/static/chunks/pages/clusters-d44859594e6f8064.js +1 -0
  138. sky/dashboard/out/_next/static/chunks/pages/config-dfb9bf07b13045f4.js +1 -0
  139. sky/dashboard/out/_next/static/chunks/pages/index-444f1804401f04ea.js +1 -0
  140. sky/dashboard/out/_next/static/chunks/pages/infra/[context]-c0b5935149902e6f.js +1 -0
  141. sky/dashboard/out/_next/static/chunks/pages/infra-aed0ea19df7cf961.js +1 -0
  142. sky/dashboard/out/_next/static/chunks/pages/jobs/[job]-5796e8d6aea291a0.js +16 -0
  143. sky/dashboard/out/_next/static/chunks/pages/jobs/pools/[pool]-6edeb7d06032adfc.js +21 -0
  144. sky/dashboard/out/_next/static/chunks/pages/jobs-479dde13399cf270.js +1 -0
  145. sky/dashboard/out/_next/static/chunks/pages/users-5ab3b907622cf0fe.js +1 -0
  146. sky/dashboard/out/_next/static/chunks/pages/volumes-b84b948ff357c43e.js +1 -0
  147. sky/dashboard/out/_next/static/chunks/pages/workspace/new-3f88a1c7e86a3f86.js +1 -0
  148. sky/dashboard/out/_next/static/chunks/pages/workspaces/[name]-c5a3eeee1c218af1.js +1 -0
  149. sky/dashboard/out/_next/static/chunks/pages/workspaces-22b23febb3e89ce1.js +1 -0
  150. sky/dashboard/out/_next/static/chunks/webpack-2679be77fc08a2f8.js +1 -0
  151. sky/dashboard/out/_next/static/css/0748ce22df867032.css +3 -0
  152. sky/dashboard/out/_next/static/zB0ed6ge_W1MDszVHhijS/_buildManifest.js +1 -0
  153. sky/dashboard/out/clusters/[cluster]/[job].html +1 -1
  154. sky/dashboard/out/clusters/[cluster].html +1 -1
  155. sky/dashboard/out/clusters.html +1 -1
  156. sky/dashboard/out/config.html +1 -0
  157. sky/dashboard/out/index.html +1 -1
  158. sky/dashboard/out/infra/[context].html +1 -0
  159. sky/dashboard/out/infra.html +1 -0
  160. sky/dashboard/out/jobs/[job].html +1 -1
  161. sky/dashboard/out/jobs/pools/[pool].html +1 -0
  162. sky/dashboard/out/jobs.html +1 -1
  163. sky/dashboard/out/users.html +1 -0
  164. sky/dashboard/out/volumes.html +1 -0
  165. sky/dashboard/out/workspace/new.html +1 -0
  166. sky/dashboard/out/workspaces/[name].html +1 -0
  167. sky/dashboard/out/workspaces.html +1 -0
  168. sky/data/data_utils.py +137 -1
  169. sky/data/mounting_utils.py +269 -84
  170. sky/data/storage.py +1451 -1807
  171. sky/data/storage_utils.py +43 -57
  172. sky/exceptions.py +132 -2
  173. sky/execution.py +206 -63
  174. sky/global_user_state.py +2374 -586
  175. sky/jobs/__init__.py +5 -0
  176. sky/jobs/client/sdk.py +242 -65
  177. sky/jobs/client/sdk_async.py +143 -0
  178. sky/jobs/constants.py +9 -8
  179. sky/jobs/controller.py +839 -277
  180. sky/jobs/file_content_utils.py +80 -0
  181. sky/jobs/log_gc.py +201 -0
  182. sky/jobs/recovery_strategy.py +398 -152
  183. sky/jobs/scheduler.py +315 -189
  184. sky/jobs/server/core.py +829 -255
  185. sky/jobs/server/server.py +156 -115
  186. sky/jobs/server/utils.py +136 -0
  187. sky/jobs/state.py +2092 -701
  188. sky/jobs/utils.py +1242 -160
  189. sky/logs/__init__.py +21 -0
  190. sky/logs/agent.py +108 -0
  191. sky/logs/aws.py +243 -0
  192. sky/logs/gcp.py +91 -0
  193. sky/metrics/__init__.py +0 -0
  194. sky/metrics/utils.py +443 -0
  195. sky/models.py +78 -1
  196. sky/optimizer.py +164 -70
  197. sky/provision/__init__.py +90 -4
  198. sky/provision/aws/config.py +147 -26
  199. sky/provision/aws/instance.py +135 -50
  200. sky/provision/azure/instance.py +10 -5
  201. sky/provision/common.py +13 -1
  202. sky/provision/cudo/cudo_machine_type.py +1 -1
  203. sky/provision/cudo/cudo_utils.py +14 -8
  204. sky/provision/cudo/cudo_wrapper.py +72 -71
  205. sky/provision/cudo/instance.py +10 -6
  206. sky/provision/do/instance.py +10 -6
  207. sky/provision/do/utils.py +4 -3
  208. sky/provision/docker_utils.py +114 -23
  209. sky/provision/fluidstack/instance.py +13 -8
  210. sky/provision/gcp/__init__.py +1 -0
  211. sky/provision/gcp/config.py +301 -19
  212. sky/provision/gcp/constants.py +218 -0
  213. sky/provision/gcp/instance.py +36 -8
  214. sky/provision/gcp/instance_utils.py +18 -4
  215. sky/provision/gcp/volume_utils.py +247 -0
  216. sky/provision/hyperbolic/__init__.py +12 -0
  217. sky/provision/hyperbolic/config.py +10 -0
  218. sky/provision/hyperbolic/instance.py +437 -0
  219. sky/provision/hyperbolic/utils.py +373 -0
  220. sky/provision/instance_setup.py +93 -14
  221. sky/provision/kubernetes/__init__.py +5 -0
  222. sky/provision/kubernetes/config.py +9 -52
  223. sky/provision/kubernetes/constants.py +17 -0
  224. sky/provision/kubernetes/instance.py +789 -247
  225. sky/provision/kubernetes/manifests/fusermount-server-daemonset.yaml +1 -2
  226. sky/provision/kubernetes/network.py +27 -17
  227. sky/provision/kubernetes/network_utils.py +40 -43
  228. sky/provision/kubernetes/utils.py +1192 -531
  229. sky/provision/kubernetes/volume.py +282 -0
  230. sky/provision/lambda_cloud/instance.py +22 -16
  231. sky/provision/nebius/constants.py +50 -0
  232. sky/provision/nebius/instance.py +19 -6
  233. sky/provision/nebius/utils.py +196 -91
  234. sky/provision/oci/instance.py +10 -5
  235. sky/provision/paperspace/instance.py +10 -7
  236. sky/provision/paperspace/utils.py +1 -1
  237. sky/provision/primeintellect/__init__.py +10 -0
  238. sky/provision/primeintellect/config.py +11 -0
  239. sky/provision/primeintellect/instance.py +454 -0
  240. sky/provision/primeintellect/utils.py +398 -0
  241. sky/provision/provisioner.py +110 -36
  242. sky/provision/runpod/__init__.py +5 -0
  243. sky/provision/runpod/instance.py +27 -6
  244. sky/provision/runpod/utils.py +51 -18
  245. sky/provision/runpod/volume.py +180 -0
  246. sky/provision/scp/__init__.py +15 -0
  247. sky/provision/scp/config.py +93 -0
  248. sky/provision/scp/instance.py +531 -0
  249. sky/provision/seeweb/__init__.py +11 -0
  250. sky/provision/seeweb/config.py +13 -0
  251. sky/provision/seeweb/instance.py +807 -0
  252. sky/provision/shadeform/__init__.py +11 -0
  253. sky/provision/shadeform/config.py +12 -0
  254. sky/provision/shadeform/instance.py +351 -0
  255. sky/provision/shadeform/shadeform_utils.py +83 -0
  256. sky/provision/ssh/__init__.py +18 -0
  257. sky/provision/vast/instance.py +13 -8
  258. sky/provision/vast/utils.py +10 -7
  259. sky/provision/vsphere/common/vim_utils.py +1 -2
  260. sky/provision/vsphere/instance.py +15 -10
  261. sky/provision/vsphere/vsphere_utils.py +9 -19
  262. sky/py.typed +0 -0
  263. sky/resources.py +844 -118
  264. sky/schemas/__init__.py +0 -0
  265. sky/schemas/api/__init__.py +0 -0
  266. sky/schemas/api/responses.py +225 -0
  267. sky/schemas/db/README +4 -0
  268. sky/schemas/db/env.py +90 -0
  269. sky/schemas/db/global_user_state/001_initial_schema.py +124 -0
  270. sky/schemas/db/global_user_state/002_add_workspace_to_cluster_history.py +35 -0
  271. sky/schemas/db/global_user_state/003_fix_initial_revision.py +61 -0
  272. sky/schemas/db/global_user_state/004_is_managed.py +34 -0
  273. sky/schemas/db/global_user_state/005_cluster_event.py +32 -0
  274. sky/schemas/db/global_user_state/006_provision_log.py +41 -0
  275. sky/schemas/db/global_user_state/007_cluster_event_request_id.py +34 -0
  276. sky/schemas/db/global_user_state/008_skylet_ssh_tunnel_metadata.py +34 -0
  277. sky/schemas/db/global_user_state/009_last_activity_and_launched_at.py +89 -0
  278. sky/schemas/db/global_user_state/010_save_ssh_key.py +66 -0
  279. sky/schemas/db/script.py.mako +28 -0
  280. sky/schemas/db/serve_state/001_initial_schema.py +67 -0
  281. sky/schemas/db/skypilot_config/001_initial_schema.py +30 -0
  282. sky/schemas/db/spot_jobs/001_initial_schema.py +97 -0
  283. sky/schemas/db/spot_jobs/002_cluster_pool.py +42 -0
  284. sky/schemas/db/spot_jobs/003_pool_hash.py +34 -0
  285. sky/schemas/db/spot_jobs/004_job_file_contents.py +42 -0
  286. sky/schemas/db/spot_jobs/005_logs_gc.py +38 -0
  287. sky/schemas/generated/__init__.py +0 -0
  288. sky/schemas/generated/autostopv1_pb2.py +36 -0
  289. sky/schemas/generated/autostopv1_pb2.pyi +43 -0
  290. sky/schemas/generated/autostopv1_pb2_grpc.py +146 -0
  291. sky/schemas/generated/jobsv1_pb2.py +86 -0
  292. sky/schemas/generated/jobsv1_pb2.pyi +254 -0
  293. sky/schemas/generated/jobsv1_pb2_grpc.py +542 -0
  294. sky/schemas/generated/managed_jobsv1_pb2.py +74 -0
  295. sky/schemas/generated/managed_jobsv1_pb2.pyi +278 -0
  296. sky/schemas/generated/managed_jobsv1_pb2_grpc.py +278 -0
  297. sky/schemas/generated/servev1_pb2.py +58 -0
  298. sky/schemas/generated/servev1_pb2.pyi +115 -0
  299. sky/schemas/generated/servev1_pb2_grpc.py +322 -0
  300. sky/serve/autoscalers.py +357 -5
  301. sky/serve/client/impl.py +310 -0
  302. sky/serve/client/sdk.py +47 -139
  303. sky/serve/client/sdk_async.py +130 -0
  304. sky/serve/constants.py +10 -8
  305. sky/serve/controller.py +64 -19
  306. sky/serve/load_balancer.py +106 -60
  307. sky/serve/load_balancing_policies.py +115 -1
  308. sky/serve/replica_managers.py +273 -162
  309. sky/serve/serve_rpc_utils.py +179 -0
  310. sky/serve/serve_state.py +554 -251
  311. sky/serve/serve_utils.py +733 -220
  312. sky/serve/server/core.py +66 -711
  313. sky/serve/server/impl.py +1093 -0
  314. sky/serve/server/server.py +21 -18
  315. sky/serve/service.py +133 -48
  316. sky/serve/service_spec.py +135 -16
  317. sky/serve/spot_placer.py +3 -0
  318. sky/server/auth/__init__.py +0 -0
  319. sky/server/auth/authn.py +50 -0
  320. sky/server/auth/loopback.py +38 -0
  321. sky/server/auth/oauth2_proxy.py +200 -0
  322. sky/server/common.py +475 -181
  323. sky/server/config.py +81 -23
  324. sky/server/constants.py +44 -6
  325. sky/server/daemons.py +229 -0
  326. sky/server/html/token_page.html +185 -0
  327. sky/server/metrics.py +160 -0
  328. sky/server/requests/executor.py +528 -138
  329. sky/server/requests/payloads.py +351 -17
  330. sky/server/requests/preconditions.py +21 -17
  331. sky/server/requests/process.py +112 -29
  332. sky/server/requests/request_names.py +120 -0
  333. sky/server/requests/requests.py +817 -224
  334. sky/server/requests/serializers/decoders.py +82 -31
  335. sky/server/requests/serializers/encoders.py +140 -22
  336. sky/server/requests/threads.py +106 -0
  337. sky/server/rest.py +417 -0
  338. sky/server/server.py +1290 -284
  339. sky/server/state.py +20 -0
  340. sky/server/stream_utils.py +345 -57
  341. sky/server/uvicorn.py +217 -3
  342. sky/server/versions.py +270 -0
  343. sky/setup_files/MANIFEST.in +5 -0
  344. sky/setup_files/alembic.ini +156 -0
  345. sky/setup_files/dependencies.py +136 -31
  346. sky/setup_files/setup.py +44 -42
  347. sky/sky_logging.py +102 -5
  348. sky/skylet/attempt_skylet.py +1 -0
  349. sky/skylet/autostop_lib.py +129 -8
  350. sky/skylet/configs.py +27 -20
  351. sky/skylet/constants.py +171 -19
  352. sky/skylet/events.py +105 -21
  353. sky/skylet/job_lib.py +335 -104
  354. sky/skylet/log_lib.py +297 -18
  355. sky/skylet/log_lib.pyi +44 -1
  356. sky/skylet/ray_patches/__init__.py +17 -3
  357. sky/skylet/ray_patches/autoscaler.py.diff +18 -0
  358. sky/skylet/ray_patches/cli.py.diff +19 -0
  359. sky/skylet/ray_patches/command_runner.py.diff +17 -0
  360. sky/skylet/ray_patches/log_monitor.py.diff +20 -0
  361. sky/skylet/ray_patches/resource_demand_scheduler.py.diff +32 -0
  362. sky/skylet/ray_patches/updater.py.diff +18 -0
  363. sky/skylet/ray_patches/worker.py.diff +41 -0
  364. sky/skylet/services.py +564 -0
  365. sky/skylet/skylet.py +63 -4
  366. sky/skylet/subprocess_daemon.py +103 -29
  367. sky/skypilot_config.py +506 -99
  368. sky/ssh_node_pools/__init__.py +1 -0
  369. sky/ssh_node_pools/core.py +135 -0
  370. sky/ssh_node_pools/server.py +233 -0
  371. sky/task.py +621 -137
  372. sky/templates/aws-ray.yml.j2 +10 -3
  373. sky/templates/azure-ray.yml.j2 +1 -1
  374. sky/templates/do-ray.yml.j2 +1 -1
  375. sky/templates/gcp-ray.yml.j2 +57 -0
  376. sky/templates/hyperbolic-ray.yml.j2 +67 -0
  377. sky/templates/jobs-controller.yaml.j2 +27 -24
  378. sky/templates/kubernetes-loadbalancer.yml.j2 +2 -0
  379. sky/templates/kubernetes-ray.yml.j2 +607 -51
  380. sky/templates/lambda-ray.yml.j2 +1 -1
  381. sky/templates/nebius-ray.yml.j2 +33 -12
  382. sky/templates/paperspace-ray.yml.j2 +1 -1
  383. sky/templates/primeintellect-ray.yml.j2 +71 -0
  384. sky/templates/runpod-ray.yml.j2 +9 -1
  385. sky/templates/scp-ray.yml.j2 +3 -50
  386. sky/templates/seeweb-ray.yml.j2 +108 -0
  387. sky/templates/shadeform-ray.yml.j2 +72 -0
  388. sky/templates/sky-serve-controller.yaml.j2 +22 -2
  389. sky/templates/websocket_proxy.py +178 -18
  390. sky/usage/usage_lib.py +18 -11
  391. sky/users/__init__.py +0 -0
  392. sky/users/model.conf +15 -0
  393. sky/users/permission.py +387 -0
  394. sky/users/rbac.py +121 -0
  395. sky/users/server.py +720 -0
  396. sky/users/token_service.py +218 -0
  397. sky/utils/accelerator_registry.py +34 -5
  398. sky/utils/admin_policy_utils.py +84 -38
  399. sky/utils/annotations.py +16 -5
  400. sky/utils/asyncio_utils.py +78 -0
  401. sky/utils/auth_utils.py +153 -0
  402. sky/utils/benchmark_utils.py +60 -0
  403. sky/utils/cli_utils/status_utils.py +159 -86
  404. sky/utils/cluster_utils.py +31 -9
  405. sky/utils/command_runner.py +354 -68
  406. sky/utils/command_runner.pyi +93 -3
  407. sky/utils/common.py +35 -8
  408. sky/utils/common_utils.py +310 -87
  409. sky/utils/config_utils.py +87 -5
  410. sky/utils/context.py +402 -0
  411. sky/utils/context_utils.py +222 -0
  412. sky/utils/controller_utils.py +264 -89
  413. sky/utils/dag_utils.py +31 -12
  414. sky/utils/db/__init__.py +0 -0
  415. sky/utils/db/db_utils.py +470 -0
  416. sky/utils/db/migration_utils.py +133 -0
  417. sky/utils/directory_utils.py +12 -0
  418. sky/utils/env_options.py +13 -0
  419. sky/utils/git.py +567 -0
  420. sky/utils/git_clone.sh +460 -0
  421. sky/utils/infra_utils.py +195 -0
  422. sky/utils/kubernetes/cleanup-tunnel.sh +62 -0
  423. sky/utils/kubernetes/config_map_utils.py +133 -0
  424. sky/utils/kubernetes/create_cluster.sh +13 -27
  425. sky/utils/kubernetes/delete_cluster.sh +10 -7
  426. sky/utils/kubernetes/deploy_remote_cluster.py +1299 -0
  427. sky/utils/kubernetes/exec_kubeconfig_converter.py +22 -31
  428. sky/utils/kubernetes/generate_kind_config.py +6 -66
  429. sky/utils/kubernetes/generate_kubeconfig.sh +4 -1
  430. sky/utils/kubernetes/gpu_labeler.py +5 -5
  431. sky/utils/kubernetes/kubernetes_deploy_utils.py +354 -47
  432. sky/utils/kubernetes/ssh-tunnel.sh +379 -0
  433. sky/utils/kubernetes/ssh_utils.py +221 -0
  434. sky/utils/kubernetes_enums.py +8 -15
  435. sky/utils/lock_events.py +94 -0
  436. sky/utils/locks.py +368 -0
  437. sky/utils/log_utils.py +300 -6
  438. sky/utils/perf_utils.py +22 -0
  439. sky/utils/resource_checker.py +298 -0
  440. sky/utils/resources_utils.py +249 -32
  441. sky/utils/rich_utils.py +213 -37
  442. sky/utils/schemas.py +905 -147
  443. sky/utils/serialize_utils.py +16 -0
  444. sky/utils/status_lib.py +10 -0
  445. sky/utils/subprocess_utils.py +38 -15
  446. sky/utils/tempstore.py +70 -0
  447. sky/utils/timeline.py +24 -52
  448. sky/utils/ux_utils.py +84 -15
  449. sky/utils/validator.py +11 -1
  450. sky/utils/volume.py +86 -0
  451. sky/utils/yaml_utils.py +111 -0
  452. sky/volumes/__init__.py +13 -0
  453. sky/volumes/client/__init__.py +0 -0
  454. sky/volumes/client/sdk.py +149 -0
  455. sky/volumes/server/__init__.py +0 -0
  456. sky/volumes/server/core.py +258 -0
  457. sky/volumes/server/server.py +122 -0
  458. sky/volumes/volume.py +212 -0
  459. sky/workspaces/__init__.py +0 -0
  460. sky/workspaces/core.py +655 -0
  461. sky/workspaces/server.py +101 -0
  462. sky/workspaces/utils.py +56 -0
  463. skypilot_nightly-1.0.0.dev20251107.dist-info/METADATA +675 -0
  464. skypilot_nightly-1.0.0.dev20251107.dist-info/RECORD +594 -0
  465. {skypilot_nightly-1.0.0.dev20250509.dist-info → skypilot_nightly-1.0.0.dev20251107.dist-info}/WHEEL +1 -1
  466. sky/benchmark/benchmark_state.py +0 -256
  467. sky/benchmark/benchmark_utils.py +0 -641
  468. sky/clouds/service_catalog/constants.py +0 -7
  469. sky/dashboard/out/_next/static/LksQgChY5izXjokL3LcEu/_buildManifest.js +0 -1
  470. sky/dashboard/out/_next/static/chunks/236-f49500b82ad5392d.js +0 -6
  471. sky/dashboard/out/_next/static/chunks/312-c3c8845990db8ffc.js +0 -15
  472. sky/dashboard/out/_next/static/chunks/37-0a572fe0dbb89c4d.js +0 -6
  473. sky/dashboard/out/_next/static/chunks/678-206dddca808e6d16.js +0 -59
  474. sky/dashboard/out/_next/static/chunks/845-0f8017370869e269.js +0 -1
  475. sky/dashboard/out/_next/static/chunks/979-7bf73a4c7cea0f5c.js +0 -1
  476. sky/dashboard/out/_next/static/chunks/fd9d1056-2821b0f0cabcd8bd.js +0 -1
  477. sky/dashboard/out/_next/static/chunks/framework-87d061ee6ed71b28.js +0 -33
  478. sky/dashboard/out/_next/static/chunks/main-app-241eb28595532291.js +0 -1
  479. sky/dashboard/out/_next/static/chunks/main-e0e2335212e72357.js +0 -1
  480. sky/dashboard/out/_next/static/chunks/pages/_app-e6b013bc3f77ad60.js +0 -1
  481. sky/dashboard/out/_next/static/chunks/pages/_error-1be831200e60c5c0.js +0 -1
  482. sky/dashboard/out/_next/static/chunks/pages/clusters/[cluster]/[job]-e15db85d0ea1fbe1.js +0 -1
  483. sky/dashboard/out/_next/static/chunks/pages/clusters/[cluster]-f383db7389368ea7.js +0 -1
  484. sky/dashboard/out/_next/static/chunks/pages/clusters-a93b93e10b8b074e.js +0 -1
  485. sky/dashboard/out/_next/static/chunks/pages/index-f9f039532ca8cbc4.js +0 -1
  486. sky/dashboard/out/_next/static/chunks/pages/jobs/[job]-03f279c6741fb48b.js +0 -1
  487. sky/dashboard/out/_next/static/chunks/pages/jobs-a75029b67aab6a2e.js +0 -1
  488. sky/dashboard/out/_next/static/chunks/webpack-830f59b8404e96b8.js +0 -1
  489. sky/dashboard/out/_next/static/css/c6933bbb2ce7f4dd.css +0 -3
  490. sky/jobs/dashboard/dashboard.py +0 -223
  491. sky/jobs/dashboard/static/favicon.ico +0 -0
  492. sky/jobs/dashboard/templates/index.html +0 -831
  493. sky/jobs/server/dashboard_utils.py +0 -69
  494. sky/skylet/providers/scp/__init__.py +0 -2
  495. sky/skylet/providers/scp/config.py +0 -149
  496. sky/skylet/providers/scp/node_provider.py +0 -578
  497. sky/templates/kubernetes-ssh-jump.yml.j2 +0 -94
  498. sky/utils/db_utils.py +0 -100
  499. sky/utils/kubernetes/deploy_remote_cluster.sh +0 -308
  500. sky/utils/kubernetes/ssh_jump_lifecycle_manager.py +0 -191
  501. skypilot_nightly-1.0.0.dev20250509.dist-info/METADATA +0 -361
  502. skypilot_nightly-1.0.0.dev20250509.dist-info/RECORD +0 -396
  503. /sky/{clouds/service_catalog → catalog}/config.py +0 -0
  504. /sky/{benchmark → catalog/data_fetchers}/__init__.py +0 -0
  505. /sky/{clouds/service_catalog → catalog}/data_fetchers/fetch_azure.py +0 -0
  506. /sky/{clouds/service_catalog → catalog}/data_fetchers/fetch_fluidstack.py +0 -0
  507. /sky/{clouds/service_catalog → catalog}/data_fetchers/fetch_ibm.py +0 -0
  508. /sky/{clouds/service_catalog/data_fetchers → client/cli}/__init__.py +0 -0
  509. /sky/dashboard/out/_next/static/{LksQgChY5izXjokL3LcEu → zB0ed6ge_W1MDszVHhijS}/_ssgManifest.js +0 -0
  510. {skypilot_nightly-1.0.0.dev20250509.dist-info → skypilot_nightly-1.0.0.dev20251107.dist-info}/entry_points.txt +0 -0
  511. {skypilot_nightly-1.0.0.dev20250509.dist-info → skypilot_nightly-1.0.0.dev20251107.dist-info}/licenses/LICENSE +0 -0
  512. {skypilot_nightly-1.0.0.dev20250509.dist-info → skypilot_nightly-1.0.0.dev20251107.dist-info}/top_level.txt +0 -0
sky/skylet/skylet.py CHANGED
@@ -1,11 +1,20 @@
1
1
  """skylet: a daemon running on the head node of a cluster."""
2
2
 
3
+ import concurrent.futures
4
+ import os
3
5
  import time
4
6
 
7
+ import grpc
8
+
5
9
  import sky
6
10
  from sky import sky_logging
11
+ from sky.schemas.generated import autostopv1_pb2_grpc
12
+ from sky.schemas.generated import jobsv1_pb2_grpc
13
+ from sky.schemas.generated import managed_jobsv1_pb2_grpc
14
+ from sky.schemas.generated import servev1_pb2_grpc
7
15
  from sky.skylet import constants
8
16
  from sky.skylet import events
17
+ from sky.skylet import services
9
18
 
10
19
  # Use the explicit logger name so that the logger is under the
11
20
  # `sky.skylet.skylet` namespace when executed directly, so as
@@ -24,12 +33,62 @@ EVENTS = [
24
33
  # This is for monitoring controller job status. If it becomes
25
34
  # unhealthy, this event will correctly update the controller
26
35
  # status to CONTROLLER_FAILED.
27
- events.ServiceUpdateEvent(),
36
+ events.ServiceUpdateEvent(pool=False),
37
+ # Status refresh for pool.
38
+ events.ServiceUpdateEvent(pool=True),
28
39
  # Report usage heartbeat every 10 minutes.
29
40
  events.UsageHeartbeatReportEvent(),
30
41
  ]
31
42
 
32
- while True:
33
- time.sleep(events.EVENT_CHECKING_INTERVAL_SECONDS)
43
+
44
+ def start_grpc_server(port: int = constants.SKYLET_GRPC_PORT) -> grpc.Server:
45
+ """Start the gRPC server."""
46
+ # This is the default value in Python 3.8 - 3.12,
47
+ # putting it here for visibility.
48
+ # TODO(kevin): Determine the optimal max number of threads.
49
+ max_workers = min(32, (os.cpu_count() or 1) + 4)
50
+ server = grpc.server(
51
+ concurrent.futures.ThreadPoolExecutor(max_workers=max_workers))
52
+
53
+ autostopv1_pb2_grpc.add_AutostopServiceServicer_to_server(
54
+ services.AutostopServiceImpl(), server)
55
+ jobsv1_pb2_grpc.add_JobsServiceServicer_to_server(
56
+ services.JobsServiceImpl(), server)
57
+ servev1_pb2_grpc.add_ServeServiceServicer_to_server(
58
+ services.ServeServiceImpl(), server)
59
+ managed_jobsv1_pb2_grpc.add_ManagedJobsServiceServicer_to_server(
60
+ services.ManagedJobsServiceImpl(), server)
61
+
62
+ listen_addr = f'127.0.0.1:{port}'
63
+ server.add_insecure_port(listen_addr)
64
+
65
+ server.start()
66
+ logger.info(f'gRPC server started on {listen_addr}')
67
+
68
+ return server
69
+
70
+
71
+ def run_event_loop():
72
+ """Run the existing event loop."""
73
+
34
74
  for event in EVENTS:
35
- event.run()
75
+ event.start()
76
+
77
+ while True:
78
+ time.sleep(events.EVENT_CHECKING_INTERVAL_SECONDS)
79
+ for event in EVENTS:
80
+ event.run()
81
+
82
+
83
+ def main():
84
+ grpc_server = start_grpc_server()
85
+ try:
86
+ run_event_loop()
87
+ except KeyboardInterrupt:
88
+ logger.info('Shutting down skylet...')
89
+ finally:
90
+ grpc_server.stop(grace=5)
91
+
92
+
93
+ if __name__ == '__main__':
94
+ main()
@@ -4,11 +4,16 @@ processes of proc_pid.
4
4
  """
5
5
  import argparse
6
6
  import os
7
+ import signal
7
8
  import sys
8
9
  import time
10
+ from typing import List, Optional
9
11
 
10
12
  import psutil
11
13
 
14
+ # Environment variable to enable kill_pg in subprocess daemon.
15
+ USE_KILL_PG_ENV_VAR = 'SKYPILOT_SUBPROCESS_DAEMON_KILL_PG'
16
+
12
17
 
13
18
  def daemonize():
14
19
  """Detaches the process from its parent process with double-forking.
@@ -38,8 +43,74 @@ def daemonize():
38
43
  # This process is now fully detached from the original parent and terminal
39
44
 
40
45
 
41
- if __name__ == '__main__':
42
- daemonize()
46
+ def get_pgid_if_leader(pid) -> Optional[int]:
47
+ """Get the process group ID of the target process if it is the leader."""
48
+ try:
49
+ pgid = os.getpgid(pid)
50
+ # Only use process group if the target process is the leader. This is
51
+ # to avoid killing the entire process group while the target process is
52
+ # just a subprocess in the group.
53
+ if pgid == pid:
54
+ print(f'Process group {pgid} is the leader.')
55
+ return pgid
56
+ return None
57
+ except Exception: # pylint: disable=broad-except
58
+ # Process group is only available in UNIX.
59
+ return None
60
+
61
+
62
+ def kill_process_group(pgid: int) -> bool:
63
+ """Kill the target process group."""
64
+ try:
65
+ print(f'Terminating process group {pgid}...')
66
+ os.killpg(pgid, signal.SIGTERM)
67
+ except Exception: # pylint: disable=broad-except
68
+ return False
69
+
70
+ # Wait 30s for the process group to exit gracefully.
71
+ time.sleep(30)
72
+
73
+ try:
74
+ print(f'Force killing process group {pgid}...')
75
+ os.killpg(pgid, signal.SIGKILL)
76
+ except Exception: # pylint: disable=broad-except
77
+ pass
78
+
79
+ return True
80
+
81
+
82
+ def kill_process_tree(process: psutil.Process,
83
+ children: List[psutil.Process]) -> bool:
84
+ """Kill the process tree of the target process."""
85
+ if process is not None:
86
+ # Kill the target process first to avoid having more children, or fail
87
+ # the process due to the children being defunct.
88
+ children = [process] + children
89
+
90
+ if not children:
91
+ sys.exit()
92
+
93
+ for child in children:
94
+ try:
95
+ child.terminate()
96
+ except psutil.NoSuchProcess:
97
+ continue
98
+
99
+ # Wait 30s for the processes to exit gracefully.
100
+ time.sleep(30)
101
+
102
+ # SIGKILL if they're still running.
103
+ for child in children:
104
+ try:
105
+ child.kill()
106
+ except psutil.NoSuchProcess:
107
+ continue
108
+
109
+ return True
110
+
111
+
112
+ def main():
113
+ # daemonize()
43
114
  parser = argparse.ArgumentParser()
44
115
  parser.add_argument('--parent-pid', type=int, required=True)
45
116
  parser.add_argument('--proc-pid', type=int, required=True)
@@ -72,37 +143,40 @@ if __name__ == '__main__':
72
143
  except (psutil.NoSuchProcess, ValueError):
73
144
  pass
74
145
 
146
+ pgid: Optional[int] = None
147
+ if os.environ.get(USE_KILL_PG_ENV_VAR) == '1':
148
+ # Use kill_pg on UNIX system if allowed to reduce the resource usage.
149
+ # Note that both implementations might leave subprocessed uncancelled:
150
+ # - kill_process_tree(default): a subprocess is able to detach itself
151
+ # from the process tree use the same technique as daemonize(). Also,
152
+ # since we refresh the process tree per second, if the subprocess is
153
+ # launched between the [last_poll, parent_die] interval, the
154
+ # subprocess will not be captured will not be killed.
155
+ # - kill_process_group: kill_pg will kill all the processed in the group
156
+ # but if a subprocess calls setpgid(0, 0) to detach itself from the
157
+ # process group (usually to daemonize itself), the subprocess will
158
+ # not be killed.
159
+ pgid = get_pgid_if_leader(process.pid)
160
+
75
161
  if process is not None and parent_process is not None:
76
162
  # Wait for either parent or target process to exit
77
163
  while process.is_running() and parent_process.is_running():
78
- try:
79
- tmp_children = process.children(recursive=True)
80
- if tmp_children:
81
- children = tmp_children
82
- except psutil.NoSuchProcess:
83
- pass
164
+ if pgid is None:
165
+ # Refresh process tree for cleanup if process group is not
166
+ # available.
167
+ try:
168
+ tmp_children = process.children(recursive=True)
169
+ if tmp_children:
170
+ children = tmp_children
171
+ except psutil.NoSuchProcess:
172
+ pass
84
173
  time.sleep(1)
85
174
 
86
- if process is not None:
87
- # Kill the target process first to avoid having more children, or fail
88
- # the process due to the children being defunct.
89
- children = [process] + children
175
+ if pgid is not None:
176
+ kill_process_group(pgid)
177
+ else:
178
+ kill_process_tree(process, children)
90
179
 
91
- if not children:
92
- sys.exit()
93
180
 
94
- for child in children:
95
- try:
96
- child.terminate()
97
- except psutil.NoSuchProcess:
98
- continue
99
-
100
- # Wait 30s for the processes to exit gracefully.
101
- time.sleep(30)
102
-
103
- # SIGKILL if they're still running.
104
- for child in children:
105
- try:
106
- child.kill()
107
- except psutil.NoSuchProcess:
108
- continue
181
+ if __name__ == '__main__':
182
+ main()