dcgm

package
v0.0.0-...-eeff1dd Latest Latest
Warning

This package is not in the latest version of its module.

 Go to latest Published: Apr 9, 2024 License: Apache-2.0 Imports: 21 Imported by: 6

Documentation

Index

Constants

View Source 
const (
	Embedded mode = iota
	Standalone
	StartHostengine
)

const for DCGM hostengine running modes: Embedded, Standalone or StartHostengine

View Source 
const (
	DCGM_FT_BINARY                 = uint('b')
	DCGM_FT_DOUBLE                 = uint('d')
	DCGM_FT_INT64                  = uint('i')
	DCGM_FT_STRING                 = uint('s')
	DCGM_FT_TIMESTAMP              = uint('t')
	DCGM_FT_INT32_BLANK            = int64(2147483632)
	DCGM_FT_INT32_NOT_FOUND        = int64(DCGM_FT_INT32_BLANK + 1)
	DCGM_FT_INT32_NOT_SUPPORTED    = int64(DCGM_FT_INT32_BLANK + 2)
	DCGM_FT_INT32_NOT_PERMISSIONED = int64(DCGM_FT_INT32_BLANK + 3)
	DCGM_FT_INT64_BLANK            = int64(9223372036854775792)
	DCGM_FT_INT64_NOT_FOUND        = int64(DCGM_FT_INT64_BLANK + 1)
	DCGM_FT_INT64_NOT_SUPPORTED    = int64(DCGM_FT_INT64_BLANK + 2)
	DCGM_FT_INT64_NOT_PERMISSIONED = int64(DCGM_FT_INT64_BLANK + 3)
	DCGM_FT_FP64_BLANK             = 140737488355328.0
	DCGM_FT_FP64_NOT_FOUND         = float64(DCGM_FT_FP64_BLANK + 1.0)
	DCGM_FT_FP64_NOT_SUPPORTED     = float64(DCGM_FT_FP64_BLANK + 2.0)
	DCGM_FT_FP64_NOT_PERMISSIONED  = float64(DCGM_FT_FP64_BLANK + 3.0)
	DCGM_FT_STR_BLANK              = "<<<NULL>>>"
	DCGM_FT_STR_NOT_FOUND          = "<<<NOT_FOUND>>>"
	DCGM_FT_STR_NOT_SUPPORTED      = "<<<NOT_SUPPORTED>>>"
	DCGM_FT_STR_NOT_PERMISSIONED   = "<<<NOT_PERM>>>"

	DCGM_FI_UNKNOWN                                 = 0
	DCGM_FI_DRIVER_VERSION                          = 1
	DCGM_FI_NVML_VERSION                            = 2
	DCGM_FI_PROCESS_NAME                            = 3
	DCGM_FI_DEV_COUNT                               = 4
	DCGM_FI_CUDA_DRIVER_VERSION                     = 5
	DCGM_FI_DEV_NAME                                = 50
	DCGM_FI_DEV_BRAND                               = 51
	DCGM_FI_DEV_NVML_INDEX                          = 52
	DCGM_FI_DEV_SERIAL                              = 53
	DCGM_FI_DEV_UUID                                = 54
	DCGM_FI_DEV_MINOR_NUMBER                        = 55
	DCGM_FI_DEV_OEM_INFOROM_VER                     = 56
	DCGM_FI_DEV_PCI_BUSID                           = 57
	DCGM_FI_DEV_PCI_COMBINED_ID                     = 58
	DCGM_FI_DEV_PCI_SUBSYS_ID                       = 59
	DCGM_FI_GPU_TOPOLOGY_PCI                        = 60
	DCGM_FI_GPU_TOPOLOGY_NVLINK                     = 61
	DCGM_FI_GPU_TOPOLOGY_AFFINITY                   = 62
	DCGM_FI_DEV_CUDA_COMPUTE_CAPABILITY             = 63
	DCGM_FI_DEV_COMPUTE_MODE                        = 65
	DCGM_FI_DEV_PERSISTENCE_MODE                    = 66
	DCGM_FI_DEV_MIG_MODE                            = 67
	DCGM_FI_DEV_CUDA_VISIBLE_DEVICES_STR            = 68
	DCGM_FI_DEV_MIG_MAX_SLICES                      = 69
	DCGM_FI_DEV_CPU_AFFINITY_0                      = 70
	DCGM_FI_DEV_CPU_AFFINITY_1                      = 71
	DCGM_FI_DEV_CPU_AFFINITY_2                      = 72
	DCGM_FI_DEV_CPU_AFFINITY_3                      = 73
	DCGM_FI_DEV_CC_MODE                             = 74
	DCGM_FI_DEV_MIG_ATTRIBUTES                      = 75
	DCGM_FI_DEV_MIG_GI_INFO                         = 76
	DCGM_FI_DEV_MIG_CI_INFO                         = 77
	DCGM_FI_DEV_ECC_INFOROM_VER                     = 80
	DCGM_FI_DEV_POWER_INFOROM_VER                   = 81
	DCGM_FI_DEV_INFOROM_IMAGE_VER                   = 82
	DCGM_FI_DEV_INFOROM_CONFIG_CHECK                = 83
	DCGM_FI_DEV_INFOROM_CONFIG_VALID                = 84
	DCGM_FI_DEV_VBIOS_VERSION                       = 85
	DCGM_FI_DEV_MEM_AFFINITY_0                      = 86
	DCGM_FI_DEV_MEM_AFFINITY_1                      = 87
	DCGM_FI_DEV_MEM_AFFINITY_2                      = 88
	DCGM_FI_DEV_MEM_AFFINITY_3                      = 89
	DCGM_FI_DEV_BAR1_TOTAL                          = 90
	DCGM_FI_SYNC_BOOST                              = 91
	DCGM_FI_DEV_BAR1_USED                           = 92
	DCGM_FI_DEV_BAR1_FREE                           = 93
	DCGM_FI_DEV_SM_CLOCK                            = 100
	DCGM_FI_DEV_MEM_CLOCK                           = 101
	DCGM_FI_DEV_VIDEO_CLOCK                         = 102
	DCGM_FI_DEV_APP_SM_CLOCK                        = 110
	DCGM_FI_DEV_APP_MEM_CLOCK                       = 111
	DCGM_FI_DEV_CLOCK_THROTTLE_REASONS              = 112
	DCGM_FI_DEV_MAX_SM_CLOCK                        = 113
	DCGM_FI_DEV_MAX_MEM_CLOCK                       = 114
	DCGM_FI_DEV_MAX_VIDEO_CLOCK                     = 115
	DCGM_FI_DEV_AUTOBOOST                           = 120
	DCGM_FI_DEV_SUPPORTED_CLOCKS                    = 130
	DCGM_FI_DEV_MEMORY_TEMP                         = 140
	DCGM_FI_DEV_GPU_TEMP                            = 150
	DCGM_FI_DEV_MEM_MAX_OP_TEMP                     = 151
	DCGM_FI_DEV_GPU_MAX_OP_TEMP                     = 152
	DCGM_FI_DEV_POWER_USAGE                         = 155
	DCGM_FI_DEV_TOTAL_ENERGY_CONSUMPTION            = 156
	DCGM_FI_DEV_POWER_USAGE_INSTANT                 = 157
	DCGM_FI_DEV_SLOWDOWN_TEMP                       = 158
	DCGM_FI_DEV_SHUTDOWN_TEMP                       = 159
	DCGM_FI_DEV_POWER_MGMT_LIMIT                    = 160
	DCGM_FI_DEV_POWER_MGMT_LIMIT_MIN                = 161
	DCGM_FI_DEV_POWER_MGMT_LIMIT_MAX                = 162
	DCGM_FI_DEV_POWER_MGMT_LIMIT_DEF                = 163
	DCGM_FI_DEV_ENFORCED_POWER_LIMIT                = 164
	DCGM_FI_DEV_PSTATE                              = 190
	DCGM_FI_DEV_FAN_SPEED                           = 191
	DCGM_FI_DEV_PCIE_TX_THROUGHPUT                  = 200
	DCGM_FI_DEV_PCIE_RX_THROUGHPUT                  = 201
	DCGM_FI_DEV_PCIE_REPLAY_COUNTER                 = 202
	DCGM_FI_DEV_GPU_UTIL                            = 203
	DCGM_FI_DEV_MEM_COPY_UTIL                       = 204
	DCGM_FI_DEV_ACCOUNTING_DATA                     = 205
	DCGM_FI_DEV_ENC_UTIL                            = 206
	DCGM_FI_DEV_DEC_UTIL                            = 207
	DCGM_FI_DEV_XID_ERRORS                          = 230
	DCGM_FI_DEV_PCIE_MAX_LINK_GEN                   = 235
	DCGM_FI_DEV_PCIE_MAX_LINK_WIDTH                 = 236
	DCGM_FI_DEV_PCIE_LINK_GEN                       = 237
	DCGM_FI_DEV_PCIE_LINK_WIDTH                     = 238
	DCGM_FI_DEV_POWER_VIOLATION                     = 240
	DCGM_FI_DEV_THERMAL_VIOLATION                   = 241
	DCGM_FI_DEV_SYNC_BOOST_VIOLATION                = 242
	DCGM_FI_DEV_BOARD_LIMIT_VIOLATION               = 243
	DCGM_FI_DEV_LOW_UTIL_VIOLATION                  = 244
	DCGM_FI_DEV_RELIABILITY_VIOLATION               = 245
	DCGM_FI_DEV_TOTAL_APP_CLOCKS_VIOLATION          = 246
	DCGM_FI_DEV_TOTAL_BASE_CLOCKS_VIOLATION         = 247
	DCGM_FI_DEV_FB_TOTAL                            = 250
	DCGM_FI_DEV_FB_FREE                             = 251
	DCGM_FI_DEV_FB_USED                             = 252
	DCGM_FI_DEV_FB_RESERVED                         = 253
	DCGM_FI_DEV_FB_USED_PERCENT                     = 254
	DCGM_FI_DEV_C2C_LINK_COUNT                      = 285
	DCGM_FI_DEV_C2C_LINK_STATUS                     = 286
	DCGM_FI_DEV_C2C_MAX_BANDWIDTH                   = 287
	DCGM_FI_DEV_ECC_CURRENT                         = 300
	DCGM_FI_DEV_ECC_PENDING                         = 301
	DCGM_FI_DEV_ECC_SBE_VOL_TOTAL                   = 310
	DCGM_FI_DEV_ECC_DBE_VOL_TOTAL                   = 311
	DCGM_FI_DEV_ECC_SBE_AGG_TOTAL                   = 312
	DCGM_FI_DEV_ECC_DBE_AGG_TOTAL                   = 313
	DCGM_FI_DEV_ECC_SBE_VOL_L1                      = 314
	DCGM_FI_DEV_ECC_DBE_VOL_L1                      = 315
	DCGM_FI_DEV_ECC_SBE_VOL_L2                      = 316
	DCGM_FI_DEV_ECC_DBE_VOL_L2                      = 317
	DCGM_FI_DEV_ECC_SBE_VOL_DEV                     = 318
	DCGM_FI_DEV_ECC_DBE_VOL_DEV                     = 319
	DCGM_FI_DEV_ECC_SBE_VOL_REG                     = 320
	DCGM_FI_DEV_ECC_DBE_VOL_REG                     = 321
	DCGM_FI_DEV_ECC_SBE_VOL_TEX                     = 322
	DCGM_FI_DEV_ECC_DBE_VOL_TEX                     = 323
	DCGM_FI_DEV_ECC_SBE_AGG_L1                      = 324
	DCGM_FI_DEV_ECC_DBE_AGG_L1                      = 325
	DCGM_FI_DEV_ECC_SBE_AGG_L2                      = 326
	DCGM_FI_DEV_ECC_DBE_AGG_L2                      = 327
	DCGM_FI_DEV_ECC_SBE_AGG_DEV                     = 328
	DCGM_FI_DEV_ECC_DBE_AGG_DEV                     = 329
	DCGM_FI_DEV_ECC_SBE_AGG_REG                     = 330
	DCGM_FI_DEV_ECC_DBE_AGG_REG                     = 331
	DCGM_FI_DEV_ECC_SBE_AGG_TEX                     = 332
	DCGM_FI_DEV_ECC_DBE_AGG_TEX                     = 333
	DCGM_FI_DEV_BANKS_REMAP_ROWS_AVAIL_MAX          = 385
	DCGM_FI_DEV_BANKS_REMAP_ROWS_AVAIL_HIGH         = 386
	DCGM_FI_DEV_BANKS_REMAP_ROWS_AVAIL_PARTIAL      = 387
	DCGM_FI_DEV_BANKS_REMAP_ROWS_AVAIL_LOW          = 388
	DCGM_FI_DEV_BANKS_REMAP_ROWS_AVAIL_NONE         = 389
	DCGM_FI_DEV_RETIRED_SBE                         = 390
	DCGM_FI_DEV_RETIRED_DBE                         = 391
	DCGM_FI_DEV_RETIRED_PENDING                     = 392
	DCGM_FI_DEV_UNCORRECTABLE_REMAPPED_ROWS         = 393
	DCGM_FI_DEV_CORRECTABLE_REMAPPED_ROWS           = 394
	DCGM_FI_DEV_ROW_REMAP_FAILURE                   = 395
	DCGM_FI_DEV_ROW_REMAP_PENDING                   = 396
	DCGM_FI_DEV_NVLINK_CRC_FLIT_ERROR_COUNT_L0      = 400
	DCGM_FI_DEV_NVLINK_CRC_FLIT_ERROR_COUNT_L1      = 401
	DCGM_FI_DEV_NVLINK_CRC_FLIT_ERROR_COUNT_L2      = 402
	DCGM_FI_DEV_NVLINK_CRC_FLIT_ERROR_COUNT_L3      = 403
	DCGM_FI_DEV_NVLINK_CRC_FLIT_ERROR_COUNT_L4      = 404
	DCGM_FI_DEV_NVLINK_CRC_FLIT_ERROR_COUNT_L5      = 405
	DCGM_FI_DEV_NVLINK_CRC_FLIT_ERROR_COUNT_TOTAL   = 409
	DCGM_FI_DEV_NVLINK_CRC_DATA_ERROR_COUNT_L0      = 410
	DCGM_FI_DEV_NVLINK_CRC_DATA_ERROR_COUNT_L1      = 411
	DCGM_FI_DEV_NVLINK_CRC_DATA_ERROR_COUNT_L2      = 412
	DCGM_FI_DEV_NVLINK_CRC_DATA_ERROR_COUNT_L3      = 413
	DCGM_FI_DEV_NVLINK_CRC_DATA_ERROR_COUNT_L4      = 414
	DCGM_FI_DEV_NVLINK_CRC_DATA_ERROR_COUNT_L5      = 415
	DCGM_FI_DEV_NVLINK_CRC_DATA_ERROR_COUNT_TOTAL   = 419
	DCGM_FI_DEV_NVLINK_REPLAY_ERROR_COUNT_L0        = 420
	DCGM_FI_DEV_NVLINK_REPLAY_ERROR_COUNT_L1        = 421
	DCGM_FI_DEV_NVLINK_REPLAY_ERROR_COUNT_L2        = 422
	DCGM_FI_DEV_NVLINK_REPLAY_ERROR_COUNT_L3        = 423
	DCGM_FI_DEV_NVLINK_REPLAY_ERROR_COUNT_L4        = 424
	DCGM_FI_DEV_NVLINK_REPLAY_ERROR_COUNT_L5        = 425
	DCGM_FI_DEV_NVLINK_REPLAY_ERROR_COUNT_TOTAL     = 429
	DCGM_FI_DEV_NVLINK_RECOVERY_ERROR_COUNT_L0      = 430
	DCGM_FI_DEV_NVLINK_RECOVERY_ERROR_COUNT_L1      = 431
	DCGM_FI_DEV_NVLINK_RECOVERY_ERROR_COUNT_L2      = 432
	DCGM_FI_DEV_NVLINK_RECOVERY_ERROR_COUNT_L3      = 433
	DCGM_FI_DEV_NVLINK_RECOVERY_ERROR_COUNT_L4      = 434
	DCGM_FI_DEV_NVLINK_RECOVERY_ERROR_COUNT_L5      = 435
	DCGM_FI_DEV_NVLINK_RECOVERY_ERROR_COUNT_TOTAL   = 439
	DCGM_FI_DEV_NVLINK_BANDWIDTH_L0                 = 440
	DCGM_FI_DEV_NVLINK_BANDWIDTH_L1                 = 441
	DCGM_FI_DEV_NVLINK_BANDWIDTH_L2                 = 442
	DCGM_FI_DEV_NVLINK_BANDWIDTH_L3                 = 443
	DCGM_FI_DEV_NVLINK_BANDWIDTH_L4                 = 444
	DCGM_FI_DEV_NVLINK_BANDWIDTH_L5                 = 445
	DCGM_FI_DEV_NVLINK_BANDWIDTH_TOTAL              = 449
	DCGM_FI_DEV_GPU_NVLINK_ERRORS                   = 450
	DCGM_FI_DEV_NVLINK_CRC_FLIT_ERROR_COUNT_L6      = 451
	DCGM_FI_DEV_NVLINK_CRC_FLIT_ERROR_COUNT_L7      = 452
	DCGM_FI_DEV_NVLINK_CRC_FLIT_ERROR_COUNT_L8      = 453
	DCGM_FI_DEV_NVLINK_CRC_FLIT_ERROR_COUNT_L9      = 454
	DCGM_FI_DEV_NVLINK_CRC_FLIT_ERROR_COUNT_L10     = 455
	DCGM_FI_DEV_NVLINK_CRC_FLIT_ERROR_COUNT_L11     = 456
	DCGM_FI_DEV_NVLINK_CRC_DATA_ERROR_COUNT_L6      = 457
	DCGM_FI_DEV_NVLINK_CRC_DATA_ERROR_COUNT_L7      = 458
	DCGM_FI_DEV_NVLINK_CRC_DATA_ERROR_COUNT_L8      = 459
	DCGM_FI_DEV_NVLINK_CRC_DATA_ERROR_COUNT_L9      = 460
	DCGM_FI_DEV_NVLINK_CRC_DATA_ERROR_COUNT_L10     = 461
	DCGM_FI_DEV_NVLINK_CRC_DATA_ERROR_COUNT_L11     = 462
	DCGM_FI_DEV_NVLINK_REPLAY_ERROR_COUNT_L6        = 463
	DCGM_FI_DEV_NVLINK_REPLAY_ERROR_COUNT_L7        = 464
	DCGM_FI_DEV_NVLINK_REPLAY_ERROR_COUNT_L8        = 465
	DCGM_FI_DEV_NVLINK_REPLAY_ERROR_COUNT_L9        = 466
	DCGM_FI_DEV_NVLINK_REPLAY_ERROR_COUNT_L10       = 467
	DCGM_FI_DEV_NVLINK_REPLAY_ERROR_COUNT_L11       = 468
	DCGM_FI_DEV_NVLINK_RECOVERY_ERROR_COUNT_L6      = 469
	DCGM_FI_DEV_NVLINK_RECOVERY_ERROR_COUNT_L7      = 470
	DCGM_FI_DEV_NVLINK_RECOVERY_ERROR_COUNT_L8      = 471
	DCGM_FI_DEV_NVLINK_RECOVERY_ERROR_COUNT_L9      = 472
	DCGM_FI_DEV_NVLINK_RECOVERY_ERROR_COUNT_L10     = 473
	DCGM_FI_DEV_NVLINK_RECOVERY_ERROR_COUNT_L11     = 474
	DCGM_FI_DEV_NVLINK_BANDWIDTH_L6                 = 475
	DCGM_FI_DEV_NVLINK_BANDWIDTH_L7                 = 476
	DCGM_FI_DEV_NVLINK_BANDWIDTH_L8                 = 477
	DCGM_FI_DEV_NVLINK_BANDWIDTH_L9                 = 478
	DCGM_FI_DEV_NVLINK_BANDWIDTH_L10                = 479
	DCGM_FI_DEV_NVLINK_BANDWIDTH_L11                = 480
	DCGM_FI_DEV_NVLINK_CRC_FLIT_ERROR_COUNT_L12     = 406
	DCGM_FI_DEV_NVLINK_CRC_FLIT_ERROR_COUNT_L13     = 407
	DCGM_FI_DEV_NVLINK_CRC_FLIT_ERROR_COUNT_L14     = 408
	DCGM_FI_DEV_NVLINK_CRC_FLIT_ERROR_COUNT_L15     = 481
	DCGM_FI_DEV_NVLINK_CRC_FLIT_ERROR_COUNT_L16     = 482
	DCGM_FI_DEV_NVLINK_CRC_FLIT_ERROR_COUNT_L17     = 483
	DCGM_FI_DEV_NVLINK_CRC_DATA_ERROR_COUNT_L12     = 416
	DCGM_FI_DEV_NVLINK_CRC_DATA_ERROR_COUNT_L13     = 417
	DCGM_FI_DEV_NVLINK_CRC_DATA_ERROR_COUNT_L14     = 418
	DCGM_FI_DEV_NVLINK_CRC_DATA_ERROR_COUNT_L15     = 484
	DCGM_FI_DEV_NVLINK_CRC_DATA_ERROR_COUNT_L16     = 485
	DCGM_FI_DEV_NVLINK_CRC_DATA_ERROR_COUNT_L17     = 486
	DCGM_FI_DEV_NVLINK_REPLAY_ERROR_COUNT_L12       = 426
	DCGM_FI_DEV_NVLINK_REPLAY_ERROR_COUNT_L13       = 427
	DCGM_FI_DEV_NVLINK_REPLAY_ERROR_COUNT_L14       = 428
	DCGM_FI_DEV_NVLINK_REPLAY_ERROR_COUNT_L15       = 487
	DCGM_FI_DEV_NVLINK_REPLAY_ERROR_COUNT_L16       = 488
	DCGM_FI_DEV_NVLINK_REPLAY_ERROR_COUNT_L17       = 489
	DCGM_FI_DEV_NVLINK_RECOVERY_ERROR_COUNT_L12     = 436
	DCGM_FI_DEV_NVLINK_RECOVERY_ERROR_COUNT_L13     = 437
	DCGM_FI_DEV_NVLINK_RECOVERY_ERROR_COUNT_L14     = 438
	DCGM_FI_DEV_NVLINK_RECOVERY_ERROR_COUNT_L15     = 491
	DCGM_FI_DEV_NVLINK_RECOVERY_ERROR_COUNT_L16     = 492
	DCGM_FI_DEV_NVLINK_RECOVERY_ERROR_COUNT_L17     = 493
	DCGM_FI_DEV_NVLINK_BANDWIDTH_L12                = 446
	DCGM_FI_DEV_NVLINK_BANDWIDTH_L13                = 447
	DCGM_FI_DEV_NVLINK_BANDWIDTH_L14                = 448
	DCGM_FI_DEV_NVLINK_BANDWIDTH_L15                = 494
	DCGM_FI_DEV_NVLINK_BANDWIDTH_L16                = 495
	DCGM_FI_DEV_NVLINK_BANDWIDTH_L17                = 496
	DCGM_FI_DEV_VIRTUAL_MODE                        = 500
	DCGM_FI_DEV_SUPPORTED_TYPE_INFO                 = 501
	DCGM_FI_DEV_CREATABLE_VGPU_TYPE_IDS             = 502
	DCGM_FI_DEV_VGPU_INSTANCE_IDS                   = 503
	DCGM_FI_DEV_VGPU_UTILIZATIONS                   = 504
	DCGM_FI_DEV_VGPU_PER_PROCESS_UTILIZATION        = 505
	DCGM_FI_DEV_ENC_STATS                           = 506
	DCGM_FI_DEV_FBC_STATS                           = 507
	DCGM_FI_DEV_FBC_SESSIONS_INFO                   = 508
	DCGM_FI_DEV_SUPPORTED_VGPU_TYPE_IDS             = 509
	DCGM_FI_DEV_VGPU_TYPE_INFO                      = 510
	DCGM_FI_DEV_VGPU_TYPE_NAME                      = 511
	DCGM_FI_DEV_VGPU_TYPE_CLASS                     = 512
	DCGM_FI_DEV_VGPU_TYPE_LICENSE                   = 513
	DCGM_FI_DEV_VGPU_VM_ID                          = 520
	DCGM_FI_DEV_VGPU_VM_NAME                        = 521
	DCGM_FI_DEV_VGPU_TYPE                           = 522
	DCGM_FI_DEV_VGPU_UUID                           = 523
	DCGM_FI_DEV_VGPU_DRIVER_VERSION                 = 524
	DCGM_FI_DEV_VGPU_MEMORY_USAGE                   = 525
	DCGM_FI_DEV_VGPU_LICENSE_STATUS                 = 526
	DCGM_FI_DEV_VGPU_FRAME_RATE_LIMIT               = 527
	DCGM_FI_DEV_VGPU_ENC_STATS                      = 528
	DCGM_FI_DEV_VGPU_ENC_SESSIONS_INFO              = 529
	DCGM_FI_DEV_VGPU_FBC_STATS                      = 530
	DCGM_FI_DEV_VGPU_FBC_SESSIONS_INFO              = 531
	DCGM_FI_DEV_VGPU_INSTANCE_LICENSE_STATE         = 532
	DCGM_FI_DEV_VGPU_PCI_ID                         = 533
	DCGM_FI_DEV_VGPU_VM_GPU_INSTANCE_ID             = 534
	DCGM_FI_INTERNAL_FIELDS_0_START                 = 600
	DCGM_FI_INTERNAL_FIELDS_0_END                   = 699
	DCGM_FI_DEV_NVSWITCH_VOLTAGE_MVOLT              = 701
	DCGM_FI_DEV_NVSWITCH_CURRENT_IDDQ               = 702
	DCGM_FI_DEV_NVSWITCH_CURRENT_IDDQ_REV           = 703
	DCGM_FI_DEV_NVSWITCH_CURRENT_IDDQ_DVDD          = 704
	DCGM_FI_DEV_NVSWITCH_POWER_VDD                  = 705
	DCGM_FI_DEV_NVSWITCH_POWER_DVDD                 = 706
	DCGM_FI_DEV_NVSWITCH_POWER_HVDD                 = 707
	DCGM_FI_DEV_NVSWITCH_LINK_THROUGHPUT_TX         = 780
	DCGM_FI_DEV_NVSWITCH_LINK_THROUGHPUT_RX         = 781
	DCGM_FI_DEV_NVSWITCH_LINK_FATAL_ERRORS          = 782
	DCGM_FI_DEV_NVSWITCH_LINK_NON_FATAL_ERRORS      = 783
	DCGM_FI_DEV_NVSWITCH_LINK_REPLAY_ERRORS         = 784
	DCGM_FI_DEV_NVSWITCH_LINK_RECOVERY_ERRORS       = 785
	DCGM_FI_DEV_NVSWITCH_LINK_FLIT_ERRORS           = 786
	DCGM_FI_DEV_NVSWITCH_LINK_CRC_ERRORS            = 787
	DCGM_FI_DEV_NVSWITCH_LINK_ECC_ERRORS            = 788
	DCGM_FI_DEV_NVSWITCH_LINK_LATENCY_LOW_VC0       = 789
	DCGM_FI_DEV_NVSWITCH_LINK_LATENCY_LOW_VC1       = 790
	DCGM_FI_DEV_NVSWITCH_LINK_LATENCY_LOW_VC2       = 791
	DCGM_FI_DEV_NVSWITCH_LINK_LATENCY_LOW_VC3       = 792
	DCGM_FI_DEV_NVSWITCH_LINK_LATENCY_MEDIUM_VC0    = 793
	DCGM_FI_DEV_NVSWITCH_LINK_LATENCY_MEDIUM_VC1    = 794
	DCGM_FI_DEV_NVSWITCH_LINK_LATENCY_MEDIUM_VC2    = 795
	DCGM_FI_DEV_NVSWITCH_LINK_LATENCY_MEDIUM_VC3    = 796
	DCGM_FI_DEV_NVSWITCH_LINK_LATENCY_HIGH_VC0      = 797
	DCGM_FI_DEV_NVSWITCH_LINK_LATENCY_HIGH_VC1      = 798
	DCGM_FI_DEV_NVSWITCH_LINK_LATENCY_HIGH_VC2      = 799
	DCGM_FI_DEV_NVSWITCH_LINK_LATENCY_HIGH_VC3      = 800
	DCGM_FI_DEV_NVSWITCH_LINK_LATENCY_PANIC_VC0     = 801
	DCGM_FI_DEV_NVSWITCH_LINK_LATENCY_PANIC_VC1     = 802
	DCGM_FI_DEV_NVSWITCH_LINK_LATENCY_PANIC_VC2     = 803
	DCGM_FI_DEV_NVSWITCH_LINK_LATENCY_PANIC_VC3     = 804
	DCGM_FI_DEV_NVSWITCH_LINK_LATENCY_COUNT_VC0     = 805
	DCGM_FI_DEV_NVSWITCH_LINK_LATENCY_COUNT_VC1     = 806
	DCGM_FI_DEV_NVSWITCH_LINK_LATENCY_COUNT_VC2     = 807
	DCGM_FI_DEV_NVSWITCH_LINK_LATENCY_COUNT_VC3     = 808
	DCGM_FI_DEV_NVSWITCH_LINK_CRC_ERRORS_LANE0      = 809
	DCGM_FI_DEV_NVSWITCH_LINK_CRC_ERRORS_LANE1      = 810
	DCGM_FI_DEV_NVSWITCH_LINK_CRC_ERRORS_LANE2      = 811
	DCGM_FI_DEV_NVSWITCH_LINK_CRC_ERRORS_LANE3      = 812
	DCGM_FI_DEV_NVSWITCH_LINK_ECC_ERRORS_LANE0      = 813
	DCGM_FI_DEV_NVSWITCH_LINK_ECC_ERRORS_LANE1      = 814
	DCGM_FI_DEV_NVSWITCH_LINK_ECC_ERRORS_LANE2      = 815
	DCGM_FI_DEV_NVSWITCH_LINK_ECC_ERRORS_LANE3      = 816
	DCGM_FI_DEV_NVSWITCH_FATAL_ERRORS               = 856
	DCGM_FI_DEV_NVSWITCH_NON_FATAL_ERRORS           = 857
	DCGM_FI_DEV_NVSWITCH_TEMPERATURE_CURRENT        = 858
	DCGM_FI_DEV_NVSWITCH_TEMPERATURE_LIMIT_SLOWDOWN = 859
	DCGM_FI_DEV_NVSWITCH_TEMPERATURE_LIMIT_SHUTDOWN = 860
	DCGM_FI_DEV_NVSWITCH_THROUGHPUT_TX              = 861
	DCGM_FI_DEV_NVSWITCH_THROUGHPUT_RX              = 862
	DCGM_FI_DEV_NVSWITCH_PHYS_ID                    = 863
	DCGM_FI_DEV_NVSWITCH_RESET_REQUIRED             = 864
	DCGM_FI_DEV_NVSWITCH_LINK_ID                    = 865
	DCGM_FI_DEV_NVSWITCH_PCIE_DOMAIN                = 866
	DCGM_FI_DEV_NVSWITCH_PCIE_BUS                   = 867
	DCGM_FI_DEV_NVSWITCH_PCIE_DEVICE                = 868
	DCGM_FI_DEV_NVSWITCH_PCIE_FUNCTION              = 869
	DCGM_FI_DEV_NVSWITCH_LINK_STATUS                = 870
	DCGM_FI_DEV_NVSWITCH_LINK_TYPE                  = 871
	DCGM_FI_DEV_NVSWITCH_LINK_REMOTE_PCIE_DOMAIN    = 872
	DCGM_FI_DEV_NVSWITCH_LINK_REMOTE_PCIE_BUS       = 873
	DCGM_FI_DEV_NVSWITCH_LINK_REMOTE_PCIE_DEVICE    = 874
	DCGM_FI_DEV_NVSWITCH_LINK_REMOTE_PCIE_FUNCTION  = 875
	DCGM_FI_DEV_NVSWITCH_LINK_DEVICE_LINK_ID        = 876
	DCGM_FI_DEV_NVSWITCH_LINK_DEVICE_LINK_SID       = 877
	DCGM_FI_DEV_NVSWITCH_LINK_DEVICE_UUID           = 878
	DCGM_FI_PROF_GR_ENGINE_ACTIVE                   = 1001
	DCGM_FI_PROF_SM_ACTIVE                          = 1002
	DCGM_FI_PROF_SM_OCCUPANCY                       = 1003
	DCGM_FI_PROF_PIPE_TENSOR_ACTIVE                 = 1004
	DCGM_FI_PROF_DRAM_ACTIVE                        = 1005
	DCGM_FI_PROF_PIPE_FP64_ACTIVE                   = 1006
	DCGM_FI_PROF_PIPE_FP32_ACTIVE                   = 1007
	DCGM_FI_PROF_PIPE_FP16_ACTIVE                   = 1008
	DCGM_FI_PROF_PCIE_TX_BYTES                      = 1009
	DCGM_FI_PROF_PCIE_RX_BYTES                      = 1010
	DCGM_FI_PROF_NVLINK_TX_BYTES                    = 1011
	DCGM_FI_PROF_NVLINK_RX_BYTES                    = 1012
	DCGM_FI_PROF_PIPE_TENSOR_IMMA_ACTIVE            = 1013
	DCGM_FI_PROF_PIPE_TENSOR_HMMA_ACTIVE            = 1014
	DCGM_FI_PROF_PIPE_TENSOR_DFMA_ACTIVE            = 1015
	DCGM_FI_PROF_PIPE_INT_ACTIVE                    = 1016
	DCGM_FI_PROF_NVDEC0_ACTIVE                      = 1017
	DCGM_FI_PROF_NVDEC1_ACTIVE                      = 1018
	DCGM_FI_PROF_NVDEC2_ACTIVE                      = 1019
	DCGM_FI_PROF_NVDEC3_ACTIVE                      = 1020
	DCGM_FI_PROF_NVDEC4_ACTIVE                      = 1021
	DCGM_FI_PROF_NVDEC5_ACTIVE                      = 1022
	DCGM_FI_PROF_NVDEC6_ACTIVE                      = 1023
	DCGM_FI_PROF_NVDEC7_ACTIVE                      = 1024
	DCGM_FI_PROF_NVJPG0_ACTIVE                      = 1025
	DCGM_FI_PROF_NVJPG1_ACTIVE                      = 1026
	DCGM_FI_PROF_NVJPG2_ACTIVE                      = 1027
	DCGM_FI_PROF_NVJPG3_ACTIVE                      = 1028
	DCGM_FI_PROF_NVJPG4_ACTIVE                      = 1029
	DCGM_FI_PROF_NVJPG5_ACTIVE                      = 1030
	DCGM_FI_PROF_NVJPG6_ACTIVE                      = 1031
	DCGM_FI_PROF_NVJPG7_ACTIVE                      = 1032
	DCGM_FI_PROF_NVOFA0_ACTIVE                      = 1033
	DCGM_FI_PROF_NVLINK_L0_TX_BYTES                 = 1040
	DCGM_FI_PROF_NVLINK_L0_RX_BYTES                 = 1041
	DCGM_FI_PROF_NVLINK_L1_TX_BYTES                 = 1042
	DCGM_FI_PROF_NVLINK_L1_RX_BYTES                 = 1043
	DCGM_FI_PROF_NVLINK_L2_TX_BYTES                 = 1044
	DCGM_FI_PROF_NVLINK_L2_RX_BYTES                 = 1045
	DCGM_FI_PROF_NVLINK_L3_TX_BYTES                 = 1046
	DCGM_FI_PROF_NVLINK_L3_RX_BYTES                 = 1047
	DCGM_FI_PROF_NVLINK_L4_TX_BYTES                 = 1048
	DCGM_FI_PROF_NVLINK_L4_RX_BYTES                 = 1049
	DCGM_FI_PROF_NVLINK_L5_TX_BYTES                 = 1050
	DCGM_FI_PROF_NVLINK_L5_RX_BYTES                 = 1051
	DCGM_FI_PROF_NVLINK_L6_TX_BYTES                 = 1052
	DCGM_FI_PROF_NVLINK_L6_RX_BYTES                 = 1053
	DCGM_FI_PROF_NVLINK_L7_TX_BYTES                 = 1054
	DCGM_FI_PROF_NVLINK_L7_RX_BYTES                 = 1055
	DCGM_FI_PROF_NVLINK_L8_TX_BYTES                 = 1056
	DCGM_FI_PROF_NVLINK_L8_RX_BYTES                 = 1057
	DCGM_FI_PROF_NVLINK_L9_TX_BYTES                 = 1058
	DCGM_FI_PROF_NVLINK_L9_RX_BYTES                 = 1059
	DCGM_FI_PROF_NVLINK_L10_TX_BYTES                = 1060
	DCGM_FI_PROF_NVLINK_L10_RX_BYTES                = 1061
	DCGM_FI_PROF_NVLINK_L11_TX_BYTES                = 1062
	DCGM_FI_PROF_NVLINK_L11_RX_BYTES                = 1063
	DCGM_FI_PROF_NVLINK_L12_TX_BYTES                = 1064
	DCGM_FI_PROF_NVLINK_L12_RX_BYTES                = 1065
	DCGM_FI_PROF_NVLINK_L13_TX_BYTES                = 1066
	DCGM_FI_PROF_NVLINK_L13_RX_BYTES                = 1067
	DCGM_FI_PROF_NVLINK_L14_TX_BYTES                = 1068
	DCGM_FI_PROF_NVLINK_L14_RX_BYTES                = 1069
	DCGM_FI_PROF_NVLINK_L15_TX_BYTES                = 1070
	DCGM_FI_PROF_NVLINK_L15_RX_BYTES                = 1071
	DCGM_FI_PROF_NVLINK_L16_TX_BYTES                = 1072
	DCGM_FI_PROF_NVLINK_L16_RX_BYTES                = 1073
	DCGM_FI_PROF_NVLINK_L17_TX_BYTES                = 1074
	DCGM_FI_PROF_NVLINK_L17_RX_BYTES                = 1075
	DCGM_FI_DEV_CPU_UTIL_TOTAL                      = 1100
	DCGM_FI_DEV_CPU_UTIL_USER                       = 1101
	DCGM_FI_DEV_CPU_UTIL_NICE                       = 1102
	DCGM_FI_DEV_CPU_UTIL_SYS                        = 1103
	DCGM_FI_DEV_CPU_UTIL_IRQ                        = 1104
	DCGM_FI_DEV_CPU_TEMP_CURRENT                    = 1110
	DCGM_FI_DEV_CPU_TEMP_WARNING                    = 1111
	DCGM_FI_DEV_CPU_TEMP_CRITICAL                   = 1112
	DCGM_FI_DEV_CPU_CLOCK_CURRENT                   = 1120
	DCGM_FI_DEV_CPU_POWER_UTIL_CURRENT              = 1130
	DCGM_FI_DEV_CPU_POWER_LIMIT                     = 1131
	DCGM_FI_DEV_CPU_VENDOR                          = 1140
	DCGM_FI_DEV_CPU_MODEL                           = 1141
	DCGM_FI_MAX_FIELDS                              = 1142

	DCGM_ST_OK                          = 0
	DCGM_ST_BADPARAM                    = -1
	DCGM_ST_GENERIC_ERROR               = -3
	DCGM_ST_MEMORY                      = -4
	DCGM_ST_NOT_CONFIGURED              = -5
	DCGM_ST_NOT_SUPPORTED               = -6
	DCGM_ST_INIT_ERROR                  = -7
	DCGM_ST_NVML_ERROR                  = -8
	DCGM_ST_PENDING                     = -9
	DCGM_ST_UNINITIALIZED               = -10
	DCGM_ST_TIMEOUT                     = -11
	DCGM_ST_VER_MISMATCH                = -12
	DCGM_ST_UNKNOWN_FIELD               = -13
	DCGM_ST_NO_DATA                     = -14
	DCGM_ST_STALE_DATA                  = -15
	DCGM_ST_NOT_WATCHED                 = -16
	DCGM_ST_NO_PERMISSION               = -17
	DCGM_ST_GPU_IS_LOST                 = -18
	DCGM_ST_RESET_REQUIRED              = -19
	DCGM_ST_FUNCTION_NOT_FOUND          = -20
	DCGM_ST_CONNECTION_NOT_VALID        = -21
	DCGM_ST_GPU_NOT_SUPPORTED           = -22
	DCGM_ST_GROUP_INCOMPATIBLE          = -23
	DCGM_ST_MAX_LIMIT                   = -24
	DCGM_ST_LIBRARY_NOT_FOUND           = -25
	DCGM_ST_DUPLICATE_KEY               = -26
	DCGM_ST_GPU_IN_SYNC_BOOST_GROUP     = -27
	DCGM_ST_GPU_NOT_IN_SYNC_BOOST_GROUP = -28
	DCGM_ST_REQUIRES_ROOT               = -29
	DCGM_ST_NVVS_ERROR                  = -30
	DCGM_ST_INSUFFICIENT_SIZE           = -31
	DCGM_ST_FIELD_UNSUPPORTED_BY_API    = -32
	DCGM_ST_MODULE_NOT_LOADED           = -33
	DCGM_ST_IN_USE                      = -34
	DCGM_ST_GROUP_IS_EMPTY              = -35
	DCGM_ST_PROFILING_NOT_SUPPORTED     = -36
	DCGM_ST_PROFILING_LIBRARY_ERROR     = -37
	DCGM_ST_PROFILING_MULTI_PASS        = -38
	DCGM_ST_DIAG_ALREADY_RUNNING        = -39
	DCGM_ST_DIAG_BAD_JSON               = -40
	DCGM_ST_DIAG_BAD_LAUNCH             = -41
	DCGM_ST_DIAG_UNUSED                 = -42
	DCGM_ST_DIAG_THRESHOLD_EXCEEDED     = -43
	DCGM_ST_INSUFFICIENT_DRIVER_VERSION = -44
	DCGM_ST_INSTANCE_NOT_FOUND          = -45
	DCGM_ST_COMPUTE_INSTANCE_NOT_FOUND  = -46
	DCGM_ST_CHILD_NOT_KILLED            = -47
	DCGM_ST_3RD_PARTY_LIBRARY_ERROR     = -48
	DCGM_ST_INSUFFICIENT_RESOURCES      = -49
	DCGM_ST_PLUGIN_EXCEPTION            = -50
	DCGM_ST_NVVS_ISOLATE_ERROR          = -51
	DCGM_ST_NVVS_BINARY_NOT_FOUND       = -52
	DCGM_ST_NVVS_KILLED                 = -53
	DCGM_ST_PAUSED                      = -54
	DCGM_ST_ALREADY_INITIALIZED         = -55
)
View Source 
const (
	MAX_NUM_CPU_CORES          uint = C.DCGM_MAX_NUM_CPU_CORES
	MAX_NUM_CPUS               uint = C.DCGM_MAX_NUM_CPUS
	CHAR_BIT                   uint = C.CHAR_BIT
	MAX_CPU_CORE_BITMASK_COUNT uint = 1024 / 8 / 8
)
View Source 
const (
	PerfStateMax     = 0
	PerfStateMin     = 15
	PerfStateUnknown = 32
)
View Source 
const (
	DiagQuick    DiagType = 1
	DiagMedium            = 2
	DiagLong              = 3
	DiagExtended          = 4
)
View Source 
const (
	MAX_NUM_DEVICES    uint = C.DCGM_MAX_NUM_DEVICES
	MAX_HIERARCHY_INFO uint = C.DCGM_MAX_HIERARCHY_INFO
)
View Source 
const (
	DbePolicy     = policyCondition("Double-bit ECC error")
	PCIePolicy    = policyCondition("PCI error")
	MaxRtPgPolicy = policyCondition("Max Retired Pages Limit")
	ThermalPolicy = policyCondition("Thermal Limit")
	PowerPolicy   = policyCondition("Power Limit")
	NvlinkPolicy  = policyCondition("Nvlink Error")
	XidPolicy     = policyCondition("XID Error")
)
View Source 
const (
	DCGM_FV_FLAG_LIVE_DATA = uint(0x00000001)
)
View Source 
const (
	DCGM_GROUP_MAX_ENTITIES int = C.DCGM_GROUP_MAX_ENTITIES
)
View Source 
const DIAG_RESULT_STRING_SIZE = 1024

Variables

View Source 
var (
	DCGM_FI = map[string]Short{}/* 423 elements not displayed */

)
View Source 
var (
	OLD_DCGM_FI = map[string]Short{
		"dcgm_sm_clock":                          100,
		"dcgm_memory_clock":                      101,
		"dcgm_memory_temp":                       140,
		"dcgm_gpu_temp":                          150,
		"dcgm_power_usage":                       155,
		"dcgm_total_energy_consumption":          156,
		"dcgm_pcie_tx_throughput":                200,
		"dcgm_pcie_rx_throughput":                201,
		"dcgm_pcie_replay_counter":               202,
		"dcgm_gpu_utilization":                   203,
		"dcgm_mem_copy_utilization":              204,
		"dcgm_enc_utilization":                   206,
		"dcgm_dec_utilization":                   207,
		"dcgm_xid_errors":                        230,
		"dcgm_power_violation":                   240,
		"dcgm_thermal_violation":                 241,
		"dcgm_sync_boost_violation":              242,
		"dcgm_board_limit_violation":             243,
		"dcgm_low_util_violation":                244,
		"dcgm_reliability_violation":             245,
		"dcgm_fb_free":                           251,
		"dcgm_fb_used":                           252,
		"dcgm_ecc_sbe_volatile_total":            310,
		"dcgm_ecc_dbe_volatile_total":            311,
		"dcgm_ecc_sbe_aggregate_total":           312,
		"dcgm_ecc_dbe_aggregate_total":           313,
		"dcgm_retired_pages_sbe":                 390,
		"dcgm_retired_pages_dbe":                 391,
		"dcgm_retired_pages_pending":             392,
		"dcgm_nvlink_flit_crc_error_count_total": 409,
		"dcgm_nvlink_data_crc_error_count_total": 419,
		"dcgm_nvlink_replay_error_count_total":   429,
		"dcgm_nvlink_recovery_error_count_total": 439,
		"dcgm_nvlink_bandwidth_total":            449,
		"dcgm_fi_prof_gr_engine_active":          1001,
		"dcgm_fi_prof_sm_active":                 1002,
		"dcgm_fi_prof_sm_occupancy":              1003,
		"dcgm_fi_prof_pipe_tensor_active":        1004,
		"dcgm_fi_prof_dram_active":               1005,
		"dcgm_fi_prof_pcie_tx_bytes":             1009,
		"dcgm_fi_prof_pcie_rx_bytes":             1010,
	}
)

Functions

func AddEntityToGroup 

func AddEntityToGroup(groupId GroupHandle, entityGroupId Field_Entity_Group, entityId uint) (err error)

func AddLinkEntityToGroup 

func AddLinkEntityToGroup(groupId GroupHandle, index uint, parentId uint) (err error)

func AddToGroup 

func AddToGroup(groupId GroupHandle, gpuId uint) (err error)

func CreateFakeEntities 

func CreateFakeEntities(entities []MigHierarchyInfo) ([]uint, error)

func DestroyGroup 

func DestroyGroup(groupId GroupHandle) (err error)

func FieldGroupDestroy 

func FieldGroupDestroy(fieldsGroup FieldHandle) (err error)

func FieldsInit 

func FieldsInit() int

func FieldsTerm 

func FieldsTerm() int

func FindFirstNonAsciiIndex 

func FindFirstNonAsciiIndex(value [4096]byte) int

func Fv2_Blob 

func Fv2_Blob(fv FieldValue_v2) [4096]byte

func Fv2_Float64 deprecated 

func Fv2_Float64(fv FieldValue_v2) float64

Deprecated: Fv2_Float64 exists for backward compatibility and should not be used. To access the int64 returned by a FieldValue_v2, use the FieldValue_v2.Float64 method.

func Fv2_Int64 deprecated 

func Fv2_Int64(fv FieldValue_v2) int64

Deprecated: Fv2_Int64 exists for backward compatibility and should not be used. To access the int64 returned by a FieldValue_v2, use the FieldValue_v2.Int64 method.

func Fv2_String 

func Fv2_String(fv FieldValue_v2) string

func GetAllDeviceCount 

func GetAllDeviceCount() (uint, error)

GetAllDeviceCount counts all GPUs on the system

func GetEntityGroupEntities 

func GetEntityGroupEntities(entityGroup Field_Entity_Group) ([]uint, error)

func GetSupportedDevices 

func GetSupportedDevices() ([]uint, error)

GetSupportedDevices returns only DCGM supported GPUs

func Init 

func Init(m mode, args ...string) (cleanup func(), err error)

Init starts DCGM, based on the user selected mode DCGM can be started in 3 differengt modes: 1. Embedded: Start hostengine within this process 2. Standalone: Connect to an already running nv-hostengine at the specified address Connection address can be passed as command line args: -connect "IP:PORT/Socket" -socket "isSocket" 3. StartHostengine: Open an Unix socket to start and connect to the nv-hostengine and terminate before exiting

func InjectFieldValue 

func InjectFieldValue(gpu uint, fieldID uint, fieldType uint, status int, ts int64, value interface{}) error

func IsInt32Blank 

func IsInt32Blank(value int) bool

func IsInt64Blank 

func IsInt64Blank(value int64) bool

func ListenForPolicyViolations 

func ListenForPolicyViolations(ctx context.Context, typ ...policyCondition) (<-chan PolicyViolation, error)

ListenForPolicyViolations sets GPU usage and error policies and notifies in case of any violations

func Shutdown 

func Shutdown() (err error)

Shutdown stops DCGM and destroy all connections

func UpdateAllFields 

func UpdateAllFields() error

func ViolationRegistration 

func ViolationRegistration(data unsafe.Pointer) int

ViolationRegistration is a go callback function for dcgmPolicyRegister() wrapped in C.violationNotify()

func WatchFieldsWithGroup 

func WatchFieldsWithGroup(fieldsGroup FieldHandle, group GroupHandle) error

func WatchFieldsWithGroupEx 

func WatchFieldsWithGroupEx(
	fieldsGroup FieldHandle, group GroupHandle, updateFreq int64, maxKeepAge float64, maxKeepSamples int32,
) error

Types

type ClockInfo 

type ClockInfo struct {
	Cores  int64 // MHz
	Memory int64 // MHz
}

type CpuHierarchyCpu_v1 

type CpuHierarchyCpu_v1 struct {
	CpuId      uint
	OwnedCores []uint64
}

type CpuHierarchy_v1 

type CpuHierarchy_v1 struct {
	Version uint
	NumCpus uint
	Cpus    [MAX_NUM_CPUS]CpuHierarchyCpu_v1
}

func GetCpuHierarchy 

func GetCpuHierarchy() (hierarchy CpuHierarchy_v1, err error)

type DbePolicyCondition 

type DbePolicyCondition struct {
	Location  string
	NumErrors uint
}

type DcgmError 

type DcgmError struct {
	Code C.dcgmReturn_t // dcgmReturn_t value of error
	// contains filtered or unexported fields
}

func (*DcgmError) Error 

func (e *DcgmError) Error() string

type DcgmStatus 

type DcgmStatus struct {
	Memory int64
	CPU    float64
}

func Introspect 

func Introspect() (DcgmStatus, error)

Introspect returns DCGM hostengine memory and CPU usage

type Device 

type Device struct {
	GPU           uint
	DCGMSupported string
	UUID          string
	Power         uint // W
	PCI           PCIInfo
	Identifiers   DeviceIdentifiers
	Topology      []P2PLink
	CPUAffinity   string
}

func GetDeviceInfo 

func GetDeviceInfo(gpuId uint) (Device, error)

GetDeviceInfo describes the given device

type DeviceHealth 

type DeviceHealth struct {
	GPU     uint
	Status  string
	Watches []SystemWatch
}

func HealthCheckByGpuId 

func HealthCheckByGpuId(gpuId uint) (DeviceHealth, error)

HealthCheckByGpuId monitors GPU health for any errors/failures/warnings

type DeviceIdentifiers 

type DeviceIdentifiers struct {
	Brand               string
	Model               string
	Serial              string
	Vbios               string
	InforomImageVersion string
	DriverVersion       string
}

type DeviceStatus 

type DeviceStatus struct {
	Power       float64 // W
	Temperature int64   // °C
	Utilization UtilizationInfo
	Memory      MemoryInfo
	Clocks      ClockInfo
	PCI         PCIStatusInfo
	Performance PerfState
	FanSpeed    int64 // %
}

func GetDeviceStatus 

func GetDeviceStatus(gpuId uint) (DeviceStatus, error)

GetDeviceStatus monitors GPU status including its power, memory and GPU utilization

type DiagResult 

type DiagResult struct {
	Status       string
	TestName     string
	TestOutput   string
	ErrorCode    uint
	ErrorMessage string
}

type DiagResults 

type DiagResults struct {
	Software []DiagResult
	PerGpu   []GpuResult
}

func RunDiag 

func RunDiag(diagType DiagType, groupId GroupHandle) (DiagResults, error)

type DiagType 

type DiagType int

type ECCErrorsInfo 

type ECCErrorsInfo struct {
	SingleBit int64
	DoubleBit int64
}

type FieldHandle 

type FieldHandle struct {
	// contains filtered or unexported fields
}

func FieldGroupCreate 

func FieldGroupCreate(fieldsGroupName string, fields []Short) (fieldsId FieldHandle, err error)

func (*FieldHandle) GetHandle 

func (f *FieldHandle) GetHandle() uintptr

func (*FieldHandle) SetHandle 

func (f *FieldHandle) SetHandle(val uintptr)

type FieldMeta 

type FieldMeta struct {
	FieldId     Short
	FieldType   byte
	Size        byte
	Tag         string
	Scope       int
	NvmlFieldId int
	EntityLevel Field_Entity_Group
}

func FieldGetById 

func FieldGetById(fieldId Short) FieldMeta

func ToFieldMeta 

func ToFieldMeta(fieldInfo C.dcgm_field_meta_p) FieldMeta

type FieldValue_v1 

type FieldValue_v1 struct {
	Version   uint
	FieldId   uint
	FieldType uint
	Status    int
	Ts        int64
	Value     [4096]byte
}

func EntityGetLatestValues 

func EntityGetLatestValues(entityGroup Field_Entity_Group, entityId uint, fields []Short) ([]FieldValue_v1, error)

func GetLatestValuesForFields 

func GetLatestValuesForFields(gpu uint, fields []Short) ([]FieldValue_v1, error)

func LinkGetLatestValues 

func LinkGetLatestValues(index uint, parentId uint, fields []Short) ([]FieldValue_v1, error)

func (FieldValue_v1) Blob 

func (fv FieldValue_v1) Blob() [4096]byte

func (FieldValue_v1) Float64 

func (fv FieldValue_v1) Float64() float64

func (FieldValue_v1) Int64 

func (fv FieldValue_v1) Int64() int64

func (FieldValue_v1) String 

func (fv FieldValue_v1) String() string

type FieldValue_v2 

type FieldValue_v2 struct {
	Version       uint
	EntityGroupId Field_Entity_Group
	EntityId      uint
	FieldId       uint
	FieldType     uint
	Status        int
	Ts            int64
	Value         [4096]byte
	StringValue   *string
}

func EntitiesGetLatestValues 

func EntitiesGetLatestValues(entities []GroupEntityPair, fields []Short, flags uint) ([]FieldValue_v2, error)

func GetValuesSince 

func GetValuesSince(GPUGroup GroupHandle, fieldGroup FieldHandle, sinceTime time.Time) ([]FieldValue_v2, time.Time, error)

GetValuesSince reads and returns field values for a specified group of entities, such as GPUs, that have been updated since a given timestamp. It allows for targeted data retrieval based on time criteria.

GPUGroup is a GroupHandle that identifies the group of entities to operate on. It can be obtained from CreateGroup for a specific group of GPUs or use GroupAllGPUs() to target all GPUs.

fieldGroup is a FieldHandle representing the group of fields for which data is requested.

sinceTime is a time.Time value representing the timestamp from which to request updated values. A zero value (time.Time{}) requests all available data.

Returns []FieldValue_v2 slice containing the requested field values, a time.Time indicating the time of the latest data retrieval, and an error if there is any issue during the operation.

func (FieldValue_v2) Blob 

func (fv FieldValue_v2) Blob() [4096]byte

func (FieldValue_v2) Float64 

func (fv FieldValue_v2) Float64() float64

func (FieldValue_v2) Int64 

func (fv FieldValue_v2) Int64() int64

func (FieldValue_v2) String 

func (fv FieldValue_v2) String() string

type Field_Entity_Group 

type Field_Entity_Group uint
const (
	FE_NONE Field_Entity_Group = iota
	FE_GPU
	FE_VGPU
	FE_SWITCH
	FE_GPU_I
	FE_GPU_CI
	FE_LINK
	FE_CPU
	FE_CPU_CORE
	FE_COUNT
)

func (Field_Entity_Group) String 

func (e Field_Entity_Group) String() string

type GpuResult 

type GpuResult struct {
	GPU         uint
	RC          uint
	DiagResults []DiagResult
}

type GroupEntityPair 

type GroupEntityPair struct {
	EntityGroupId Field_Entity_Group
	EntityId      uint
}

type GroupHandle 

type GroupHandle struct {
	// contains filtered or unexported fields
}

func CreateGroup 

func CreateGroup(groupName string) (goGroupId GroupHandle, err error)

func GroupAllGPUs 

func GroupAllGPUs() GroupHandle

func NewDefaultGroup 

func NewDefaultGroup(groupName string) (GroupHandle, error)

func WatchFields 

func WatchFields(gpuId uint, fieldsGroup FieldHandle, groupName string) (groupId GroupHandle, err error)

func WatchPidFields 

func WatchPidFields() (GroupHandle, error)

WatchPidFields lets DCGM start recording stats for GPU process It needs to be called before calling GetProcessInfo

func WatchPidFieldsEx 

func WatchPidFieldsEx(updateFreq, maxKeepAge time.Duration, maxKeepSamples int, gpus ...uint) (GroupHandle, error)

WatchPidFieldsEx is the same as WatchPidFields, but allows for modifying the update frequency, max samples, max sample age, and the GPUs on which to enable watches.

func (*GroupHandle) GetHandle 

func (g *GroupHandle) GetHandle() uintptr

func (*GroupHandle) SetHandle 

func (g *GroupHandle) SetHandle(val uintptr)
type Link_State uint
const (
	LS_NOT_SUPPORTED Link_State = iota // Link is unsupported (Default for GPUs)
	LS_DISABLED                        // Link is supported but disabled (Default for NvSwitches)
	LS_DOWN                            // Link link is down (inactive)
	LS_UP                              // Link link is up (active)
)

type MemoryInfo 

type MemoryInfo struct {
	GlobalUsed int64
	ECCErrors  ECCErrorsInfo
}

type MetricGroup 

type MetricGroup struct {
	Major    uint
	Minor    uint
	FieldIds []uint
}

func GetSupportedMetricGroups 

func GetSupportedMetricGroups(gpuId uint) ([]MetricGroup, error)

Get all of the profiling metric groups for a given GPU group.

type MigEntityInfo 

type MigEntityInfo struct {
	GpuUuid               string
	NvmlGpuIndex          uint
	NvmlInstanceId        uint
	NvmlComputeInstanceId uint
	NvmlMigProfileId      uint
	NvmlProfileSlices     uint
}

type MigHierarchyInfo 

type MigHierarchyInfo struct {
	Entity       GroupEntityPair
	Parent       GroupEntityPair
	SliceProfile MigProfile
}

type MigHierarchyInfo_v2 

type MigHierarchyInfo_v2 struct {
	Entity GroupEntityPair
	Parent GroupEntityPair
	Info   MigEntityInfo
}

type MigHierarchy_v2 

type MigHierarchy_v2 struct {
	Version    uint
	Count      uint
	EntityList [C.DCGM_MAX_HIERARCHY_INFO]MigHierarchyInfo_v2
}

func GetGpuInstanceHierarchy 

func GetGpuInstanceHierarchy() (hierarchy MigHierarchy_v2, err error)

type MigProfile 

type MigProfile int
const (
	MigProfileNone                      MigProfile = 0  /*!< No profile (for GPUs) */
	MigProfileGPUInstanceSlice1         MigProfile = 1  /*!< GPU instance slice 1 */
	MigProfileGPUInstanceSlice2         MigProfile = 2  /*!< GPU instance slice 2 */
	MigProfileGPUInstanceSlice3         MigProfile = 3  /*!< GPU instance slice 3 */
	MigProfileGPUInstanceSlice4         MigProfile = 4  /*!< GPU instance slice 4 */
	MigProfileGPUInstanceSlice7         MigProfile = 5  /*!< GPU instance slice 7 */
	MigProfileGPUInstanceSlice8         MigProfile = 6  /*!< GPU instance slice 8 */
	MigProfileGPUInstanceSlice6         MigProfile = 7  /*!< GPU instance slice 6 */
	MigProfileGPUInstanceSlice1Rev1     MigProfile = 8  /*!< GPU instance slice 1 revision 1 */
	MigProfileGPUInstanceSlice2Rev1     MigProfile = 9  /*!< GPU instance slice 2 revision 1 */
	MigProfileGPUInstanceSlice1Rev2     MigProfile = 10 /*!< GPU instance slice 1 revision 2 */
	MigProfileComputeInstanceSlice1     MigProfile = 30 /*!< compute instance slice 1 */
	MigProfileComputeInstanceSlice2     MigProfile = 31 /*!< compute instance slice 2 */
	MigProfileComputeInstanceSlice3     MigProfile = 32 /*!< compute instance slice 3 */
	MigProfileComputeInstanceSlice4     MigProfile = 33 /*!< compute instance slice 4*/
	MigProfileComputeInstanceSlice7     MigProfile = 34 /*!< compute instance slice 7 */
	MigProfileComputeInstanceSlice8     MigProfile = 35 /*!< compute instance slice 8 */
	MigProfileComputeInstanceSlice6     MigProfile = 36 /*!< compute instance slice 6 */
	MigProfileComputeInstanceSlice1Rev1 MigProfile = 37 /*!< compute instance slice 1 revision 1 */
)

type NvLinkStatus 

type NvLinkStatus struct {
	ParentId   uint
	ParentType Field_Entity_Group
	State      Link_State
	Index      uint
}

func GetNvLinkLinkStatus 

func GetNvLinkLinkStatus() ([]NvLinkStatus, error)

type NvlinkPolicyCondition 

type NvlinkPolicyCondition struct {
	FieldId uint16
	Counter uint
}
type P2PLink struct {
	GPU   uint
	BusID string
	Link  P2PLinkType
}

func GetDeviceTopology 

func GetDeviceTopology(gpuId uint) ([]P2PLink, error)

GetDeviceTopology returns device topology corresponding to the gpuId

type P2PLinkType 

type P2PLinkType uint
const (
	P2PLinkUnknown P2PLinkType = iota
	P2PLinkCrossCPU
	P2PLinkSameCPU
	P2PLinkHostBridge
	P2PLinkMultiSwitch
	P2PLinkSingleSwitch
	P2PLinkSameBoard
	SingleNVLINKLink
	TwoNVLINKLinks
	ThreeNVLINKLinks
	FourNVLINKLinks
)

func (P2PLinkType) PCIPaths 

func (l P2PLinkType) PCIPaths() string

type PCIInfo 

type PCIInfo struct {
	BusID     string
	BAR1      uint  // MB
	FBTotal   uint  // MB
	Bandwidth int64 // MB/s
}

type PCIStatusInfo 

type PCIStatusInfo struct {
	BAR1Used   int64 // MB
	Throughput PCIThroughputInfo
	FBUsed     int64
}

type PCIThroughputInfo 

type PCIThroughputInfo struct {
	Rx      int64 // MB
	Tx      int64 // MB
	Replays int64
}

type PciPolicyCondition 

type PciPolicyCondition struct {
	ReplayCounter uint
}

type PerfState 

type PerfState uint

func (PerfState) String 

func (p PerfState) String() string

type PolicyViolation 

type PolicyViolation struct {
	Condition policyCondition
	Timestamp time.Time
	Data      interface{}
}

type PowerPolicyCondition 

type PowerPolicyCondition struct {
	PowerViolation uint
}

type ProcessInfo 

type ProcessInfo struct {
	GPU                uint
	PID                uint
	Name               string
	ProcessUtilization ProcessUtilInfo
	PCI                PCIStatusInfo
	Memory             MemoryInfo
	GpuUtilization     UtilizationInfo
	Clocks             ClockInfo
	Violations         ViolationTime
	XIDErrors          XIDErrorInfo
}

func GetProcessInfo 

func GetProcessInfo(group GroupHandle, pid uint) ([]ProcessInfo, error)

GetProcessInfo provides detailed per GPU stats for this process

type ProcessUtilInfo 

type ProcessUtilInfo struct {
	StartTime      Time
	EndTime        Time
	EnergyConsumed *uint64 // Joules
	SmUtil         *float64
	MemUtil        *float64
}

type RetiredPagesPolicyCondition 

type RetiredPagesPolicyCondition struct {
	SbePages uint
	DbePages uint
}

type Short 

type Short C.ushort

type SystemWatch 

type SystemWatch struct {
	Type   string
	Status string
	Error  string
}

type ThermalPolicyCondition 

type ThermalPolicyCondition struct {
	ThermalViolation uint
}

type Time 

type Time uint64

func (Time) String 

func (t Time) String() string

type UtilizationInfo 

type UtilizationInfo struct {
	GPU     int64 // %
	Memory  int64 // %
	Encoder int64 // %
	Decoder int64 // %
}

type ViolationTime 

type ViolationTime struct {
	Power          *uint64
	Thermal        *uint64
	Reliability    *uint64
	BoardLimit     *uint64
	LowUtilization *uint64
	SyncBoost      *uint64
}

ViolationTime measures amount of time (in ms) GPU was at reduced clocks

type XIDErrorInfo 

type XIDErrorInfo struct {
	NumErrors int
	Timestamp []uint64
}

type XidPolicyCondition 

type XidPolicyCondition struct {
	ErrNum uint
}

Source Files

View all Source files

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL
go.dev uses cookies from Google to deliver and enhance the quality of its services and to analyze traffic. Learn more.