1# OpenBMC platform telemetry 2 3Author: 4 Piotr Matuszczak <piotr.matuszczak@intel.com> 5 6Other contributors: 7 Pawel Rapkiewicz <pawel.rapkiewicz@intel.com> <pawelr>, 8 Kamil Kowalski <kamil.kowalski@intel.com> 9 10Created: 11 2019-08-07 12 13## Problem Description 14The BMC on server platform gathers lots of telemetry data, which has to 15be exposed in clean, human readable and standardized format. This document 16focuses on telemetry over the Redfish, since it is standard API 17for platform manageability. 18 19## Background and References 20* OpenBMC platform telemetry shall leverage DMTF's [Redfish Telemetry Model][1] 21for exposing platform telemetry over the network. 22* OpenBMC platform telemetry shall leverage the 23[OpenBMC sensors architecture implementation][2]. 24* OpenBMC platform telemetry shall implement a service, called Telemetry to deal 25with metrics report and trigger management. This service is described later in 26this document. 27* Although we use the [hwmon][3] to gather readings from physical sensors, this 28architecture does not depend on it, because the Telemetry service component 29relies on the [OpenBMC D-Bus sensors][2]. 30 31 32## Requirements 33* [OpenBMC D-Bus sensors][2] support. This is also design limitation, since the 34Telemetry service requires telemetry sources to be implemented as D-Bus sensors. 35 36 37## Proposed Design 38Redfish Telemetry Model shall implement Telemetry Service with the following 39collection resources: 40* Metric Definitions - contains the metadata for metrics (unit, accuracy, etc.) 41* Metric Report Definitions - defines how metric report shall be created 42(which metrics it shall contain, how often it shall be generated etc.) 43* Metric Reports - contains actual metric reports containing telemetry data 44generated according to the Metric Report Definitions 45* Metric Triggers - contains thresholds and actions that apply to specific 46metrics 47 48OpenBMC telemetry architecture is shown on the diagram below. 49 50```ascii 51 +--------------+ +----------------+ +-----------------+ 52 |hwmon| | |Dbus sensors| | |Telemetry| | 53 +-----/ | +------------/ | +---------/ | 54 | +--filesystem---> | | | 55 | | | | | | 56 +--------------+ +--------^-------+ +--------^--------+ 57 | | 58 | | 59<------------------------------------------v-----^--DBus----------v-----------> 60 | 61 | 62+-------+---------------------------------------------------------------------+ 63|bmcweb | | | 64+-------/ | | 65| | | 66| +--------+-------------------------------------v--------------------------+ | 67| |Redfish | | | 68| +--------/ +---------+-------+ | | 69| | |Existing | | | | 70| | +------------------------------------------------+ |Redfish | | | | 71| | |Telemetry Service| | |resources| | | | 72| | +----------------+/ | +---------/ | | | 73| | | +----------+ +-----------+ +-------------+ | | +---------+ | | | 74| | | | Metric | | Metric | |Metric report| | | | Redfish | | | | 75| | | | triggers | |definitions| |definitions <---------+ sensors | | | | 76| | | | | | | | | | | | | | | | 77| | | +----+-----+ +-----+-----+ +------+------+ | | +---------+ | | | 78| | | | | | | | | | | 79| | | | | | | | | | | 80| | | | | | | | | | | 81| | | | +-----v-----+ | | | | | | 82| | | | | Metric | | | | | | | 83| | | +--------> report <---------+ | | | | | 84| | | | | | | | | | 85| | | +-----------+ | | | | | 86| | | | | | | | 87| | +------------------------------------------------+ +-----------------+ | | 88| | | | 89| +-------------------------------------------------------------------------+ | 90| | 91+-----------------------------------------------------------------------------+ 92``` 93 94The telemetry service component is a part of Redfish and implements the DMTF's 95[Redfish Telemetry Model][1]. Metric report definitions uses Redfish sensors 96URIs for metric report creation. Those sensors are also used to get 97URI->D-Bus sensor mapping. Redfish Telemetry Service acts as presentation 98layer for the telemetry, while Telemetry service is responsible for gathering 99metrics from D-Bus sensors and exposing them as D-Bus objects. Telemetry 100service supports different monitoring modes (periodic, on change and on demand) 101along with aggregated operations: 102* SINGLE - current reading value 103* AVERAGE - average value over defined time period 104* MAX - max reading value during defined time period 105* MIN - min reading value during defined time period 106* SUM - sum of reading values over defined time period 107 108The time period for calculating aggregated metric is taken from the Redfish 109Metric Report Definition resource for each sensor's metric. 110 111Telemetry service supports creating and managing metric report, which may 112contain single or multiple metrics from sensors. This metric report is mapped 113to Metric Report for the Redfish Telemetry Service. 114 115The diagram below shows the flows for creation and update of metric report. 116 117```ascii 118+----+ +------+ +---------+ +-------+ 119|User| |bmcweb| |Telemetry| | D-Bus | 120+-+--+ +--+---+ +----+----+ |Sensors| 121 | | | +---+---+ 122 | | | | 123+-----------------------------------------------------------------------------+ 124|Metric report definition flow| | | | 125+-----------------------------+ | | | 126| | | | | | 127| | | | | | 128| | POST request | | | | 129| | with metric | | | | 130| | report | | | | 131| | definition | | | | 132| +--------------------> Invoke AddReport | Register for D-Bus | | 133| | | method on D-Bus | sensors | | 134| | +-----------------------> PropertiesChanged | | 135| | | | signals | | 136| | | +--------------------------> | 137| | | |--------------------------> | 138| | | +--------------------------> | 139| | | | | | 140| | HTTP response | +-+Create Report | | 141| | code 201 with | Return created | |D-Bus object | | 142| | Metric Report | Report D-Bus path <-+ | | 143| | Definition's URI <-----------------------+ | | 144| <--------------------+ | | | 145| | | | | | 146| | | | | | 147+-----------------------------------------------------------------------------+ 148 | | | | 149+-----------------------------------------------------------------------------+ 150|Periodic metric report update flow| | | | 151+----------------------------------+ +-+Metric report | | 152| | | | |timer triggers | | 153| | | <-+report update | | 154| | | | | | 155+----------------------------------Optional-----------------------------------+ 156| | | | | | 157| | Send report as SSE or push-style event | | | 158| | using Redfish Event Service (not shown | | | 159| | here) if configured to do so. | | | 160| <--------------------------------------------+ | | 161| | | | | | 162+-----------------------------------------------------------------------------+ 163| | GET on Metric | | | | 164| | Report URI | | Sensor's Properties- | | 165| +--------------------> | Changed signal | | 166| | +-+Map report's URI <--------------------------+ | 167| | | |to D-Bus path | | | 168| | <-+ | +----------------------+ | | 169| | | Invoke GetAll method | |Note that sensor's | | | 170| | | on report D-Bus | |PropertiesChanged | | | 171| | | object | |signal is asynchronous| | | 172| | +-----------------------> |to metric report timer| | | 173| | | | |This timer is the only| | | 174| | Return metric | Return report data | |thing that triggers | | | 175| | report in JSON <-----------------------+ |metric report update | | | 176| | format | | +----------------------+ | | 177| <--------------------+ | | | 178| | | | | | 179+-----------------------------------------------------------------------------+ 180 | | | | 181+-----------------------------------------------------------------------------+ 182|On change metric report update flow| | Sensor's Properties- | | 183+-----------------------------------+ | Changed signal | | 184| | | <--------------------------+ | 185| | | | | | 186| | | +-+Sensor's signal | | 187| | | | |triggers report | | 188| | | <-+update | | 189| | | | | | 190+----------------------------------Optional-----------------------------------+ 191| | | | | | 192| | Send report as SSE or push-style event | | | 193| | using Redfish Event Service (not shown | | | 194| | here) if configured to do so. | | | 195| <--------------------------------------------+ | | 196| | | | | | 197+-----------------------------------------------------------------------------+ 198| | GET on Metric | | | | 199| | Report URI | | | | 200| +--------------------> | | | 201| | +-+Map report's URI | | | 202| | | |to D-Bus path | +----------------------+ | | 203| | <-+ | |Note that sensor's | | | 204| | | Invoke GetAll method | |PropertiesChanged | | | 205| | | on report D-Bus | |signal triggers the | | | 206| | | object | |report update. It is | | | 207| | +-----------------------> |sufficient that the | | | 208| | | | |signal from only one | | | 209| | Return metric | Return report data | |sensor triggers report| | | 210| | report in JSON <-----------------------+ |update. | | | 211| | format | | +----------------------+ | | 212| <--------------------+ | | | 213| | | | | | 214+-----------------------------------------------------------------------------+ 215 | | | | 216+-+--------------------+------------------------------------------------------+ 217|On demand metric report update flow| | | | 218+-+--------------------+------------+ | | | 219| | | | | | 220| | GET on Metric | | | | 221| | Report URI | | | | 222| +--------------------> | | | 223| | +-+Map report's URI | | | 224| | | |to D-Bus path | | | 225| | <-+ | | | 226| | | | | | 227| | | Invoke the Update | | | 228| | | method for report | | | 229| | | D-Bus object | | | 230| | +-----------------------> | | 231| | | +-+Update method triggers | | 232| | | | |report to be updated | | 233| | | | |with the latest known | | 234| | | | |sensor's readings. | | 235| | | | |No additional sensor | | 236| | | <-+readings are performed. | | 237+----------------------------------Optional-----------------------------------+ 238| | | | | | 239| | Send report as SSE or push-style event | | | 240| | using Redfish Event Service (not shown | | | 241| | here) if configured to do so. | | | 242| <--------------------------------------------+ | | 243| | | | | | 244+-----------------------------------------------------------------------------+ 245| | | Update method call | | | 246| | | result | | | 247| | <-----------------------+ | | 248| | | | | | 249| | | Invoke GetAll method | | | 250| | | on report D-Bus | | | 251| | | object | | | 252| | +-----------------------> | | 253| | | | | | 254| | Return metric | Return report data | | | 255| | report in JSON <-----------------------+ | | 256| | format | | | | 257| <--------------------+ | | | 258| | | | | | 259+-----------------------------------------------------------------------------+ 260 | | | | 261``` 262 263The Redfish implementation in bmcweb is stateless, thus it is not able to 264store metric reports. All operations on metric reports shall be done in 265the Telemetry service. Sending metric report as SSE or push-style events 266shall be done via the [Redfish Event Service][6]. It is marked as optional 267because metric report does not have to be configured for pushing its data 268through the event. 269 270In case of on demand metric report update, Telemetry service performs no 271additional sensor readings because it already has the latest values, since 272they are updated on PropertiesChanged signal from the D-Bus sensors. 273 274**Telemetry service on [D-Bus][4]** 275 276Telemetry service exposes specific interfaces on D-Bus. One of them will be 277used for reading report management. The second one will be used for triggers 278management. 279 280**Reading report management** 281 282The reading report management D-Bus object: 283 284```ascii 285xyz.openbmc_project.Telemetry.ReportManager 286/xyz/openbmc_project/Telemetry/Reports 287``` 288The `ReportManager` implements D-Bus interface 289[`xyz.openbmc_project.Telemetry.ReportManager`][8] for report management. The 290interface is described in the phosphor-dbus-interfaces. This interface 291implements `AddReport` method, which is used to create a metric report. The 292report may contain a single or multiple sensor readings. The way how the report 293will be stored by the BMC is defined by one of this method's parameters. 294The `ReportManager` object implements property that stores the maximum number 295of reports supported simultaneously. 296 297The `AddReport` method returns the path to the newly created report object. 298The report object implements the [`xyz.openbmc_project.Object.Delete`][10] 299and [`xyz.openbmc_project.Telemetry.Report`][9] interfaces. The [`Delete`][10] 300interface is defined to add support for removing Report object, while the 301[`Report`][9] interface implements methods and properties for Report 302management along with properties containing telemetry readings. Each report 303object contains the timestamp of its last update. The report object 304contains an array of structures containing reading with its metadata and 305timestamp of last update of this metric. Each report has also the property 306that stores update interval (for periodically updated reports). 307 308**Trigger management** 309 310The trigger management D-Bus object: 311 312```ascii 313xyz.openbmc_project.Telemetry.TriggerManager 314/xyz/openbmc_project/Telemetry/Triggers 315``` 316The `TriggerManager` supports the 317`xyz.openbmc_project.Telemetry.TriggerManager` interface, which implements 318the `AddTrigger` method. This method shall be used to create new trigger for 319the certain metric. The method's parameters allow to define the type of metric 320for which trigger is set (discrete or numeric). Depend on this setting, this 321method accepts different set of trigger parameters. 322 323For discrete metric type, trigger parameters contain: 324 325| Field | Type | Description | 326|-------|------|-------------| 327| TriggerCondition | enum | Discrete trigger condition: <br> "Changed" - trigger occurs when value of metric has changed; <br> "Specified" - trigger occurs when value of metric becomes one of the values listed in the discrete triggers. | 328| DiscreteTriggers | array of structures | Array of discrete trigger structures. | 329 330Member of DiscreteTriggers array: 331 332| Field | Type | Description | 333|-------|------|-------------| 334| TriggerId| string | Unique trigger Id | 335| Severity | enum | Severity: <br> "OK" - normal<br> "Warning" - requires attention<br> "Critical" - requires immediate attention | 336| Value | variant | Value of discrete metric, that constitutes a trigger event. | 337| DwellTime | uint64 | Time in milliseconds that a trigger occurrence persists before the action defined for this trigger is performed. | 338 339For numeric metric type, trigger parameters contain numeric thresholds. 340Numeric thresholds structure shall contain up to 4 thresholds: upper warning, upper critical, lower warning and lower critical. Thus it will contain up to 4 structures shown below: 341 342| Field | Type | Description | 343|-------|------|-------------| 344| ThresholdType | enum | Numeric trigger type: <br> "UpperCritical" - reading is above normal range and requires immediate attention<br>"UpperWarning" - reading is above normal range and may require attention<br>"LowerCritical" - reading is below normal range and requires immediate attention<br>"LowerWarning" - reading is below normal range and may require attention| 345| DwellTime | uint64 | Time in milliseconds that a trigger occurrence persists before the action defined for this trigger is performed. | 346| Activation | enum | Indicates direction of crossing the threshold value that trigger the threshold's action:<br> "Increasing" - trigger action when reading is changing from below to above the threshold's value<br> "Decreasing" - trigger action when reading is changing from above to below the threshold's value<br> "Either" - trigger action when reading is crossing the threshold's value in either direction described above | 347| ThresholdValue | variant | Value of reading that will trigger the threshold | 348 349The `AddTrigger` method also allows to define the specific action when trigger 350is activated. Upon the trigger activation, three possible actions are allowed, 351logging event to log service, sending event via event service and triggering 352the metric report update. 353 354In order to assign trigger to specific metric, the metric parameter is defined. 355Its structure contains the following data: 356 357| Field | Type | Description | 358|-------|------|-------------| 359| SensorPath | object path | D-Bus path to sensor, for which trigger is defined. | 360| MetricId | string | Contains unique metric id, that can be mapped to Redfish MetricId. | 361| ReportPath | object path | D-Bus path to Telemetry's reading report which update shall be triggered when trigger condition occurs. This is optional and shall be filled when trigger's action is set to update metric report. | 362 363The `AddTrigger` method also allows to set trigger's persistency (whether 364trigger shall be stored in the BMC's non-volatile memory). 365 366The `AddTrigger` method returns: 367```ascii 368String for created trigger - ie. '/xyz/openbmc_project/Telemetry/Triggers/{Domain}/{Name}' 369``` 370Such created trigger implements the `xyz.openbmc_project.Object.Delete` 371and the `xyz.openbmc_project.Telemetry.Trigger` interfaces. Each trigger 372object contains read-only information about metric type, for which it was 373created (discrete or numeric). This information determines which triggers 374are stored within trigger object. 375 376If trigger is defined for discrete metric type, than it contains trigger 377information that looks like this: 378 379| Type | Description | 380|------|-------------| 381| enum | Discrete trigger condition: <br> "Changed" - trigger occurs when value of metric has changed;<br>"Specified" - trigger occurs when value of metric becomes one of the values listed in the discrete triggers. | 382| array of structures | Array of discrete trigger structures. | 383 384Discrete trigger structure: 385 386| Type | Description | 387|------|-------------| 388| string | Unique trigger Id | 389| enum | Severity: <br> "OK" - normal<br>"Warning" - requires attention<br> "Critical" - requires immediate attention | 390| variant | Value of discrete metric, that constitutes a trigger event. | 391| uint64 | Time in milliseconds that a trigger occurrence persists before the action defined in the ```ActionType``` is performed. | 392 393If trigger is defined for numeric metric type, than it contains information 394about numeric triggers that is an array of 4 structures presented below: 395 396| Type | Description | 397|------|-------------| 398| enum | Numeric trigger type: <br> "UpperWarning"<br>"UpperWarning"<br>"LowerCritical"<br>"LowerWarning"| 399| uint64 | Time in milliseconds that a trigger occurrence persists before the action defined in the ```ActionType``` is performed. | 400| enum | Indicates direction of crossing the threshold value that trigger the threshold's action:<br> "Increasing"<br>"Decreasing"<br>"Either" | 401| variant | Value of reading that will trigger the threshold | 402 403The trigger object also contains information about reading, for which trigger 404was defined. It is in a form of structure consisting of three fields. 405 406| Field type | Description | 407|------------|--------------| 408| object path | D-Bus path of sensor. This is a path to the sensor, for which reading trigger is defined. | 409| string | Unique metric Id | 410| object path | D-Bus path of existing reading report. This is required when trigger's action is to update metric report. This path shall point to existing reading report within the Telemetry service. | 411 412**Trigger operations** 413 414Triggers support three types of operation: Log, Event and Update. For each, 415there is a different way of proceeding. 416 4171. For action Log, the event shall 418be logged to the system journal. In this case the Telemetry service writes 419data to system journal using libjournal. The Redfish log service shall then 420retrieve the data by reading system journal. All is shown on the diagram below. 421 422```ascii 423+---------------------------+ 424|bmcweb| | +----------------------+ 425+------/ +-----------+-+ | |Telemetry| | 426| |Redfish | | | +---------/ | 427| |log service| | | | | 428| +-----------/ | | | | 429| | | | | | 430| | | | | | 431| +------^------+ | +-----------+----------+ 432+---------------------------+ | 433 | | 434 +----collect----+ event 435 journal entry | (write to journal) 436 | | 437 +------------------------------------+ | 438 |systemd| | | | 439 +-------/ +----------+ +---+------+ | | 440 | |journal| | |libjournal| | | 441 | +-------/ <--> <-------+ 442 | | | +----------+ | 443 | | | | 444 | | | | 445 | +----------+ | 446 | | 447 +------------------------------------+ 448``` 4492. For action Event, the Telemetry service shall send event using the 450[Redfish Event Service][6] either as push-style event or SSE. 451 4523. For action Update, the Telemetry service will trigger the update of reading 453report pointed by it's D-Bus path contained in trigger object properties. The 454update shall cause the reading report's D-Bus object to emit property change 455signal. This will cause Redfish Metric Report to be streamed out if it was 456configured to do so. 457 458**Redfish Telemetry Service API** 459 460Redfish Telemetry Service shall support 2019.1 Redfish schemas for telemetry 461resources. Metric report definitions determines which metrics are to be include 462in metric report. Metric definition is assigned to particular metric type and it 463describes how the metric should be interpreted. The following resource schemas 464shall be supported: 465 466- TelemetryService 1.1.2 467- MetricDefinition 1.0.3 468- MetricReportDefinition 1.3.0 469- MetricReport 1.2.0 470- Triggers 1.1.1 471 472The following diagram shows relations between these resources. 473 474```ascii 475 +----------------------------------------------------------------------------+ 476 | Service root | 477 +----------------------------------+-------------------------------+---------+ 478 | | 479 | | 480 | | 481 +----------------------------------v-----------------+ +----------v---------+ 482 | | |Chassis | 483 | Telemetry Service | | | 484 | | | | 485 | | | +---------------+ | 486 +---------+--------------+------------------+--------+ | | | | 487 | | | | | Chassis 1 | | 488 | | | | | | | 489 | | | | +---------+-----+ | 490 | | | | | | 491+----------v--+ +--------v----+ +----------v-----+ +--------------------+ 492|Triggers | |Metric | |Metric report | | 493| | |definition | | | | 494| | | +---------+ | | | Reads | 495| +---------+ | | |Reading | | | +-----------+ | ReadingVolts +--v------+ 496| | | | | |Volts <------+ +------------------> | 497| |Trigger 1| | | +---------+ | | | Metric | | | | 498| | | | | | | | report 1 | | Reads | Power | 499| | | | | +---------+ | | | | | PowerConsumed | | 500| | | | | | | | | | | | Watts | | 501| +--+---+--+ | | |Power <------+ +------------------> | 502| | | | | |Consumed | | | +-----^-----+ | +----^----+ 503| | | | | |Watts | | | | | | 504| | | | | +---------+ | | | | | 505| | | | | | | | | | 506+-------------+ +-------------+ +----------------+ | 507 | | | | 508 | | Triggers report update | | 509 | | (when applicable) | | 510 | +--------------------------------+ | 511 | | 512 | Monitors PowerConsumedWatts to check | 513 | whether trigger value is exceeded | 514 +------------------------------------------------------------------+ 515``` 516 517The diagram shows the relations between Redfish resources. Metric report is 518defined to be generated periodically, on demand or on change. Each metric in the 519Metric Report contains the URI to its metric definition and Redfish sensor, 520which reading value is presented. Nevertheless, under this presentation layer, 521Telemetry is gathering D-Bus sensors readings and exposing them 522in reading reports over D-Bus for the Telemetry Service. Each D-Bus sensor 523is mapped to Redfish sensor. 524 525Below examples of Redfish resources for the Telemetry Service are shown. 526 527The Telemetry Service Redfish resource example: 528 529```json 530{ 531 "@odata.type": "#TelemetryService.v1_1_2.TelemetryService", 532 "Id": "TelemetryService", 533 "Name": "Telemetry Service", 534 "Status": { 535 "State": "Enabled", 536 "Health": "OK" 537 }, 538 "MinCollectionInterval": "T00:00:10s", 539 "SupportedCollectionFunctions": [], 540 "MaxReports": <max_no_of_reports>, 541 "MetricDefinitions": { 542 "@odata.id": "/redfish/v1/TelemetryService/MetricDefinitions" 543 }, 544 "MetricReportDefinitions": { 545 "@odata.id": "/redfish/v1/TelemetryService/MetricReportDefinitions" 546 }, 547 "MetricReports": { 548 "@odata.id": "/redfish/v1/TelemetryService/MetricReports" 549 }, 550 "Triggers": { 551 "@odata.id": "/redfish/v1/TelemetryService/Triggers" 552 }, 553 "LogService": { 554 "@odata.id": "/redfish/v1/Managers/<manager_name>/LogServices/Journal" 555 }, 556 "@odata.context": "/redfish/v1/$metadata#TelemetryService", 557 "@odata.id": "/redfish/v1/TelemetryService" 558} 559``` 560 561Sample metric report definition: 562 563```json 564{ 565 "@odata.type": "#MetricReportDefinition.v1_3_0.MetricReportDefinition", 566 "Id": "SampleMetric", 567 "Name": "Sample Metric Report Definition", 568 "MetricReportDefinitionType": "Periodic", 569 "Schedule": { 570 "RecurrenceInterval": "T00:00:10" 571 }, 572 "ReportActions": [ 573 "LogToMetricReportsCollection" 574 ], 575 "ReportUpdates": "Overwrite", 576 "MetricReport": { 577 "@odata.id": "/redfish/v1/TelemetryService/MetricReports/SampleMetric" 578 }, 579 "Status": { 580 "State": "Enabled" 581 }, 582 "Metrics": [ 583 { 584 "MetricId": "Test", 585 "MetricProperties": [ 586 "/redfish/v1/Chassis/NC_Baseboard/Power#/PowerControl/0/PowerConsumedWatts" 587 ] 588 } 589 ], 590 "@odata.context": "/redfish/v1/$metadata#MetricReportDefinition.MetricReportDefinition", 591 "@odata.id": "/redfish/v1/TelemetryService/MetricReportDefinitions/PlatformPowerUsage" 592} 593``` 594 595Sample metric report: 596 597```json 598{ 599 "@odata.type": "#MetricReport.v1_2_0.MetricReport", 600 "Id": "SampleMetric", 601 "Name": "Sample Metric Report", 602 "ReportSequence": "0", 603 "MetricReportDefinition": { 604 "@odata.id": "/redfish/v1/TelemetryService/MetricReportDefinitions/SampleMetric" 605 }, 606 "MetricValues": [ 607 { 608 "MetricDefinition": { 609 "@odata.id": "/redfish/v1/TelemetryService/MetricDefinitions/SampleMetricDefinition" 610 }, 611 "MetricId": "Test", 612 "MetricValue": "100", 613 "Timestamp": "2016-11-08T12:25:00-05:00", 614 "MetricProperty": "/redfish/v1/Chassis/Tray_1/Power#/0/PowerConsumedWatts" 615 } 616 ], 617 "@odata.context": "/redfish/v1/$metadata#MetricReport.MetricReport", 618 "@odata.id": "/redfish/v1/TelemetryService/MetricReports/AvgPlatformPowerUsage" 619} 620``` 621 622Sample trigger, that will trigger metric report update: 623 624```json 625{ 626 "@odata.type": "#Triggers.v1_1_1.Triggers", 627 "Id": "SampleTrigger", 628 "Name": "Sample Trigger", 629 "MetricType": "Numeric", 630 "Links": { 631 "MetricReportDefinitions": [ 632 "/redfish/v1/TelemetryService/MetricReportDefinitions/SampleMetric" 633 ] 634 }, 635 "Status": { 636 "State": "Enabled" 637 }, 638 "TriggerActions": [ 639 "RedfishMetricReport" 640 ], 641 "NumericThresholds": { 642 "UpperCritical": { 643 "Reading": 50, 644 "Activation": "Increasing", 645 "DwellTime": "PT0.001S" 646 }, 647 "UpperWarning": { 648 "Reading": 48.1, 649 "Activation": "Increasing", 650 "DwellTime": "PT0.004S" 651 } 652 }, 653 "MetricProperties": [ 654 "/redfish/v1/Chassis/Tray_1/Power#/0/PowerConsumedWatts" 655 ], 656 "@odata.context": "/redfish/v1/$metadata#Triggers.Triggers", 657 "@odata.id": "/redfish/v1/TelemetryService/Triggers/PlatformPowerCapTriggers" 658} 659``` 660 661**Performance tests** 662 663Performance test were conducted on the AST2500 system with 64 MB flash and 664512 MB RAM. Flash consumption by the Telemetry service is 197.5 kB. The 665runtime statistics are shown in the table below. The reading report is 666mapped into single Metric Report. The runtime data is collected for the 667Telemetry component only. All reports was created with 668```xyz.openbmc_project.Telemetry.Metric.OnChange``` property to 669maximize the workload. In the configuration with 50 reports and 50 sensors 670it is about 200 new readings per second, generating 200 reading reports 671per second. The table shows CPU usage and memory usage. The VSZ is the amount 672of memory mapped into the address space of the process. It includes pages 673backed by the process' executable file and shared libraries, its heap and 674stack, as well as anything else it has mapped. 675 676 677| Telemetry service state | VSZ | %VSZ | %CPU | 678|--------------------------------------------------|------|------|------| 679| Idle (0 reports, 0 sensors) |5188 B| 1% | 0% | 680| 1 report, 1 sensor |5188 B| 1% | 1% | 681| 2 reports, 1 sensor |5188 B| 1% | 1% | 682| 2 reports, 2 sensors (1 sensor per report) |5188 B| 1% | 1% | 683| 1 report, 10 sensors |5188 B| 1% | 1% | 684| 10 reports, 10 sensors (same for each report) |5320 B| 1% | 1-2% | 685| 2 reports, 20 sensors (10 per report) |5188 B| 1% | 1% | 686| 30 reports, 30 sensors (10 per report) |5444 B| 1% | 5-9% | 687| 50 reports, 50 sensors (10 per report) |5572 B| 1% |11-14%| 688 689The last two configurations use 10 sensors per reading report, which gives 6903 or 5 distinctive configurations. Each such configuration is used to 691create 10 reading reports to obtain the desired amount of 30 or 50 reading 692reports. 693 694In this architecture reading report is created every time when Redfish 695Metric Report Definition is posted (creating new Metric Report). 696 697## Alternatives Considered 698The [framework based on collectd/librrd][5] was considered as alternate design. 699Although it seems to be versatile and scalable solution, it has some drawbacks 700from our point of view: 701* Collectd's footprint in the minimal working configuration is around 2.6 MB, 702while available space for the OpenBMC is limited to 64 MB. 703* In this design, librrd is used to store metrics on the BMC's non-volatile 704storage, which may be an issue, when lots of metrics are captured and stored 705to OpenBMC's limited storage space. Also flash wear-out issue may occur, when 706metrics are captured frequently (like once per second). 707* Telemetry service is directly compatible with Redfish Telemetry Service API, 708which means, that Telemetry's reading reports can be directly mapped to Redfish 709Metric Reports. 710* Telemetry service unifies the way how the BMC's telemetry is exposed over 711the Redfish and may be used with multiple front-ends, thus there is no problem 712to add support telemetry over IPMI or any other API. 713 714Since this design assumes flexibility and modularity, there is no obstacles to 715use collectd in cooperation with Telemetry. The one of possible configurations 716is shown on the diagram below. 717 718```ascii 719 +-----------------+ +-----------------+ 720 | D-Bus sensors | | Telemetry | 721 +--------^--------+ +--------^--------+ 722 | | 723 | | 724 | | 725<--------^--v-----------D-Bus--------v-^----------> 726 | | 727 | | 728 | | 729 +-------v------------+ +----------v--------+ 730 | collectd metrics | | | 731 | exposed as D-Bus | | bmcweb | 732 | sensors | | (with Redfish | 733 +---------^----------+ | Telemetry | 734 | | Service) | 735 | | | 736 +------+-------+ +-------------------+ 737 | | 738 | collectd | 739 | | 740 +--------------+ 741``` 742Here collectd is used as the source of some set of metrics. It exposes them 743as the D-Bus sensors, which can easily be consumed either by the bmcweb and 744Telemetry service without any changes in their D-Bus interfaces. In such 745configuration Telemetry service provides metric reports and triggers 746management. 747 748Other possible configuration is to use collectd without the Telemetry service, 749but in such case, collectd does not provide metric reports and triggers support 750compatible with the Redfish. In such case, Redfish Telemetry Service won't be 751supported or metric reports and triggers support has to be provided by the 752collectd. 753 754## Impacts 755This design impacts the architecture of the bmcweb component, since it adds 756the Redfish Telemetry Service implementation as a component for the existing 757Redfish API implementation. 758 759## Testing 760This is the very high-level description of the proposed set of tests. 761Testing shall be done on three basic levels: 762* Unit tests 763* Functional tests 764* Performance tests 765 766**Unit tests** 767 768The Telemetry's code shall be covered by the unit tests. The preferred 769framework is the [GTest/GMock][7]. The unit tests shall be ran before code 770change is to be committed to make sure, that nothing is broken in existing 771functionality. Also, when new code is introduced, a new set of unit tests shall 772be committed with it according to test-driven development principle. Unit tests 773shall be also carefully reviewed. 774 775**Functional tests** 776 777Functional tests will be divided into two steps. 778 779First step is for testing the Telemetry metric reports management. Test scenario 780shall contain creating metric report by POSTing proper metric report definition, 781reading metric report (using GET on proper URI) and deleting the metric report. 782The required configuration for such test is D-Bus sensors (at least some of 783them) and bmcweb with Redfish Telemetry Service implemented. The tests shall be 784performed on real hardware. For ease of metric testing, dummy D-Bus sensors may 785be provided to provide specifically prepared metrics. This configuration shall 786also enable testing aggregated operations (MIN, MAX, SUM, AVG). 787 788Second step is to test triggers and events generation. This will require also 789Event Service to be implemented along with Log Service. Tests shall cover all 790scenarios with sending metric report as an event, triggering metric report 791update and logging events. 792 793**Performance tests** 794 795Performance tests shall be done using full OpenBMC configuration with all 796the required set of features. The tests shall create a lot of metric reports 797(up to maximum number) along with all possible triggers. Measurements shall 798cover the periodic metric report jitter, delays in event logging or sending, 799BMC's CPU utilization and the performance impact on other services. 800 801[1]: https://www.dmtf.org/sites/default/files/standards/documents/DSP2051_0.1.0a.zip 802[2]: https://github.com/openbmc/docs/blob/master/architecture/sensor-architecture.md 803[3]: https://www.kernel.org/doc/Documentation/hwmon/ 804[4]: https://www.freedesktop.org/wiki/Software/dbus/ 805[5]: https://gerrit.openbmc-project.xyz/c/openbmc/docs/+/22257 806[6]: https://gerrit.openbmc-project.xyz/c/openbmc/docs/+/24749 807[7]: https://github.com/google/googletest 808[8]: https://github.com/openbmc/phosphor-dbus-interfaces/blob/master/yaml/xyz/openbmc_project/Telemetry/ReportManager.interface.yaml 809[9]: https://github.com/openbmc/phosphor-dbus-interfaces/blob/master/yaml/xyz/openbmc_project/Telemetry/Report.interface.yaml 810[10]: https://github.com/openbmc/phosphor-dbus-interfaces/blob/master/yaml/xyz/openbmc_project/Object/Delete.interface.yaml 811