1 Detailed architecture and interface specification
2 =================================================
4 This section describes a detailed implementation plan, which is based on the
5 high level architecture introduced in Section 3. Section 5.1 describes the
6 functional blocks of the Doctor architecture, which is followed by a high level
7 message flow in Section 5.2. Section 5.3 provides a mapping of selected existing
8 open source components to the building blocks of the Doctor architecture.
9 Thereby, the selection of components is based on their maturity and the gap
10 analysis executed in Section 4. Sections 5.4 and 5.5 detail the specification of
11 the related northbound interface and the related information elements. Finally,
12 Section 5.6 provides a first set of blueprints to address selected gaps required
13 for the realization functionalities of the Doctor project.
20 This section introduces the functional blocks to form the VIM. OpenStack was
21 selected as the candidate for implementation. Inside the VIM, 4 different
22 building blocks are defined (see :numref:`figure6`).
24 .. figure:: images/figure6.png
33 The Monitor module has the responsibility for monitoring the virtualized
34 infrastructure. There are already many existing tools and services (e.g. Zabbix)
35 to monitor different aspects of hardware and software resources which can be
36 used for this purpose.
41 The Inspector module has the ability a) to receive various failure notifications
42 regarding physical resource(s) from Monitor module(s), b) to find the affected
43 virtual resource(s) by querying the resource map in the Controller, and c) to
44 update the state of the virtual resource (and physical resource).
46 The Inspector has drivers for different types of events and resources to
47 integrate any type of Monitor and Controller modules. It also uses a failure
48 policy database to decide on the failure selection and aggregation from raw
49 events. This failure policy database is configured by the Administrator.
51 The reason for separation of the Inspector and Controller modules is to make the
52 Controller focus on simple operations by avoiding a tight integration of various
53 health check mechanisms into the Controller.
58 The Controller is responsible for maintaining the resource map (i.e. the mapping
59 from physical resources to virtual resources), accepting update requests for the
60 resource state(s) (exposing as provider API), and sending all failure events
61 regarding virtual resources to the Notifier. Optionally, the Controller has the
62 ability to force the state of a given physical resource to down in the resource
63 mapping when it receives failure notifications from the Inspector for that
64 given physical resource.
65 The Controller also re-calculates the capacity of the NVFI when receiving a
66 failure notification for a physical resource.
68 In a real-world deployment, the VIM may have several controllers, one for each
69 resource type, such as Nova, Neutron and Cinder in OpenStack. Each controller
70 maintains a database of virtual and physical resources which shall be the master
71 source for resource information inside the VIM.
76 The focus of the Notifier is on selecting and aggregating failure events
77 received from the controller based on policies mandated by the Consumer.
78 Therefore, it allows the Consumer to subscribe for alarms regarding virtual
79 resources using a method such as API endpoint. After receiving a fault
80 event from a Controller, it will notify the fault to the Consumer by referring
81 to the alarm configuration which was defined by the Consumer earlier on.
83 To reduce complexity of the Controller, it is a good approach for the
84 Controllers to emit all notifications without any filtering mechanism and have
85 another service (i.e. Notifier) handle those notifications properly. This is the
86 general philosophy of notifications in OpenStack. Note that a fault message
87 consumed by the Notifier is different from the fault message received by the
88 Inspector; the former message is related to virtual resources which are visible
89 to users with relevant ownership, whereas the latter is related to raw devices
90 or small entities which should be handled with an administrator privilege.
92 The northbound interface between the Notifier and the Consumer/Administrator is
93 specified in :ref:`impl_nbi`.
101 The detailed work flow for fault management is as follows (see also :numref:`figure7`):
103 1. Request to subscribe to monitor specific virtual resources. A query filter
104 can be used to narrow down the alarms the Consumer wants to be informed
106 2. Each subscription request is acknowledged with a subscribe response message.
107 The response message contains information about the subscribed virtual
108 resources, in particular if a subscribed virtual resource is in "alarm"
110 3. The NFVI sends monitoring events for resources the VIM has been subscribed
111 to. Note: this subscription message exchange between the VIM and NFVI is not
112 shown in this message flow.
113 4. Event correlation, fault detection and aggregation in VIM.
114 5. Database lookup to find the virtual resources affected by the detected fault.
115 6. Fault notification to Consumer.
116 7. The Consumer switches to standby configuration (STBY)
117 8. Instructions to VIM requesting certain actions to be performed on the
118 affected resources, for example migrate/update/terminate specific
119 resource(s). After reception of such instructions, the VIM is executing the
120 requested action, e.g. it will migrate or terminate a virtual resource.
122 a. Query request from Consumer to VIM to get information about the current
123 status of a resource.
124 b. Response to the query request with information about the current status of
125 the queried resource. In case the resource is in "fault" state, information
126 about the related fault(s) is returned.
128 In order to allow for quick reaction to failures, the time interval between
129 fault detection in step 3 and the corresponding recovery actions in step 7 and 8
130 shall be less than 1 second.
132 .. figure:: images/figure7.png
136 Fault management work flow
138 .. figure:: images/figure8.png
142 Fault management scenario
144 :numref:`figure8` shows a more detailed message flow (Steps 4 to 6) between
145 the 4 building blocks introduced in :ref:`impl_fb`.
147 4. The Monitor observed a fault in the NFVI and reports the raw fault to the
149 The Inspector filters and aggregates the faults using pre-configured
153 a) The Inspector queries the Resource Map to find the virtual resources
154 affected by the raw fault in the NFVI.
155 b) The Inspector updates the state of the affected virtual resources in the
157 c) The Controller observes a change of the virtual resource state and informs
158 the Notifier about the state change and the related alarm(s).
159 Alternatively, the Inspector may directly inform the Notifier about it.
161 6. The Notifier is performing another filtering and aggregation of the changes
162 and alarms based on the pre-configured alarm configuration. Finally, a fault
163 notification is sent to northbound to the Consumer.
168 The detailed work flow for NFVI maintenance is shown in :numref:`figure9`
169 and has the following steps. Note that steps 1, 2, and 5 to 8a in the NFVI
170 maintenance work flow are very similar to the steps in the fault management work
171 flow and share a similar implementation plan in Release 1.
173 1. Subscribe to fault/maintenance notifications.
174 2. Response to subscribe request.
175 3. Maintenance trigger received from administrator.
176 4. VIM switches NFVI resources to "maintenance" state. This, e.g., means they
177 should not be used for further allocation/migration requests
178 5. Database lookup to find the virtual resources affected by the detected
179 maintenance operation.
180 6. Maintenance notification to Consumer.
181 7. The Consumer switches to standby configuration (STBY)
182 8. Instructions from Consumer to VIM requesting certain recovery actions to be
183 performed (step 7a). After reception of such instructions, the VIM is
184 executing the requested action in order to empty the physical resources (step
186 9. Maintenance response from VIM to inform the Administrator that the physical
187 machines have been emptied (or the operation resulted in an error state).
188 10. Administrator is coordinating and executing the maintenance operation/work
191 A) Query request from Administrator to VIM to get information about the
192 current state of a resource.
193 B) Response to the query request with information about the current state of
194 the queried resource(s). In case the resource is in "maintenance" state,
195 information about the related maintenance operation is returned.
197 .. figure:: images/figure9.png
201 NFVI maintenance work flow
203 .. figure:: images/figure10.png
207 NFVI Maintenance implementation plan
209 :numref:`figure10` shows a more detailed message flow (Steps 4 to 6)
210 between the 4 building blocks introduced in Section 5.1..
212 3. The Administrator is sending a StateChange request to the Controller residing
214 4. The Controller queries the Resource Map to find the virtual resources
215 affected by the planned maintenance operation.
218 a) The Controller updates the state of the affected virtual resources in the
219 Resource Map database.
221 b) The Controller informs the Notifier about the virtual resources that will
222 be affected by the maintenance operation.
224 6. A maintenance notification is sent to northbound to the Consumer.
228 9. The Controller informs the Administrator after the physical resources have
233 Implementation plan for OPNFV Release 1
234 ---------------------------------------
239 :numref:`figure11` shows the implementation plan based on OpenStack and
240 related components as planned for Release 1. Hereby, the Monitor can be realized
241 by Zabbix. The Controller is realized by OpenStack Nova [NOVA]_, Neutron
242 [NEUT]_, and Cinder [CIND]_ for compute, network, and storage,
243 respectively. The Inspector can be realized by Monasca [MONA]_ or a simple
244 script querying Nova in order to map between physical and virtual resources. The
245 Notifier will be realized by Ceilometer [CEIL]_ receiving failure events
246 on its notification bus.
248 :numref:`figure12` shows the inner-workings of Ceilometer. After receiving
249 an "event" on its notification bus, first a notification agent will grab the
250 event and send a "notification" to the Collector. The collector writes the
251 notifications received to the Ceilometer databases.
253 In the existing Ceilometer implementation, an alarm evaluator is periodically
254 polling those databases through the APIs provided. If it finds new alarms, it
255 will evaluate them based on the pre-defined alarm configuration, and depending
256 on the configuration, it will hand a message to the Alarm Notifier, which in
257 turn will send the alarm message northbound to the Consumer. :numref:`figure12`
258 also shows an optimized work flow for Ceilometer with the goal to
259 reduce the delay for fault notifications to the Consumer. The approach is to
260 implement a new notification agent (called "publisher" in Ceilometer
261 terminology) which is directly sending the alarm through the "Notification Bus"
262 to a new "Notification-driven Alarm Evaluator (NAE)" (see Sections 5.6.2 and
263 5.6.3), thereby bypassing the Collector and avoiding the additional delay of the
264 existing polling-based alarm evaluator. The NAE is similar to the OpenStack
265 "Alarm Evaluator", but is triggered by incoming notifications instead of
266 periodically polling the OpenStack "Alarms" database for new alarms. The
267 Ceilometer "Alarms" database can hold three states: "normal", "insufficient
268 data", and "fired". It is representing a persistent alarm database. In order to
269 realize the Doctor requirements, we need to define new "meters" in the database
272 .. figure:: images/figure11.png
276 Implementation plan in OpenStack (OPNFV Release 1 ”Arno”)
279 .. figure:: images/figure12.png
283 Implementation plan in Ceilometer architecture
289 For NFVI Maintenance, a quite similar implementation plan exists. Instead of a
290 raw fault being observed by the Monitor, the Administrator is sending a
291 Maintenance Request through the northbound interface towards the Controller
292 residing in the VIM. Similar to the Fault Management use case, the Controller
293 (in our case OpenStack Nova) will send a maintenance event to the Notifier (i.e.
294 Ceilometer in our implementation). Within Ceilometer, the same workflow as
295 described in the previous section applies. In addition, the Controller(s) will
296 take appropriate actions to evacuate the physical machines in order to prepare
297 them for the planned maintenance operation. After the physical machines are
298 emptied, the Controller will inform the Administrator that it can initiate the
304 This section introduces all attributes and information elements used in the
305 messages exchange on the northbound interfaces between the VIM and the VNFO and
308 Note: The information elements will be aligned with current work in ETSI NFV IFA
312 Simple information elements:
314 * SubscriptionID: identifies a subscription to receive fault or maintenance
316 * NotificationID: identifies a fault or maintenance notification.
317 * VirtualResourceID (Identifier): identifies a virtual resource affected by a
318 fault or a maintenance action of the underlying physical resource.
319 * PhysicalResourceID (Identifier): identifies a physical resource affected by a
320 fault or maintenance action.
321 * VirtualResourceState (String): state of a virtual resource, e.g. "normal",
322 "maintenance", "down", "error".
323 * PhysicalResourceState (String): state of a physical resource, e.g. "normal",
324 "maintenance", "down", "error".
325 * VirtualResourceType (String): type of the virtual resource, e.g. "virtual
326 machine", "virtual memory", "virtual storage", "virtual CPU", or "virtual
328 * FaultID (Identifier): identifies the related fault in the underlying physical
329 resource. This can be used to correlate different fault notifications caused
330 by the same fault in the physical resource.
331 * FaultType (String): Type of the fault. The allowed values for this parameter
332 depend on the type of the related physical resource. For example, a resource
333 of type "compute hardware" may have faults of type "CPU failure", "memory
334 failure", "network card failure", etc.
335 * Severity (Integer): value expressing the severity of the fault. The higher the
336 value, the more severe the fault.
337 * MinSeverity (Integer): value used in filter information elements. Only faults
338 with a severity higher than the MinSeverity value will be notified to the
340 * EventTime (Datetime): Time when the fault was observed.
341 * EventStartTime and EventEndTime (Datetime): Datetime range that can be used in
342 a FaultQueryFilter to narrow down the faults to be queried.
343 * ProbableCause: information about the probable cause of the fault.
344 * CorrelatedFaultID (Integer): list of other faults correlated to this fault.
345 * isRootCause (Boolean): Parameter indicating if this fault is the root for
346 other correlated faults. If TRUE, then the faults listed in the parameter
347 CorrelatedFaultID are caused by this fault.
348 * FaultDetails (Key-value pair): provides additional information about the
349 fault, e.g. information about the threshold, monitored attributes, indication
350 of the trend of the monitored parameter.
351 * FirmwareVersion (String): current version of the firmware of a physical
353 * HypervisorVersion (String): current version of a hypervisor.
354 * ZoneID (Identifier): Identifier of the resource zone. A resource zone is the
355 logical separation of physical and software resources in an NFVI deployment
356 for physical isolation, redundancy, or administrative designation.
357 * Metadata (Key-Value-Pairs): provides additional information of a physical
358 resource in maintenance/error state.
360 Complex information elements (see also UML diagrams in :numref:`figure13`
361 and :numref:`figure14`):
363 * VirtualResourceInfoClass:
365 + VirtualResourceID [1] (Identifier)
366 + VirtualResourceState [1] (String)
367 + Faults [0..*] (FaultClass): For each resource, all faults
368 including detailed information about the faults are provided.
370 * FaultClass: The parameters of the FaultClass are partially based on ETSI TS
371 132 111-2 (V12.1.0) [*]_, which is specifying fault management in 3GPP, in
372 particular describing the information elements used for alarm notifications.
374 - FaultID [1] (Identifier)
376 - Severity [1] (Integer)
377 - EventTime [1] (Datetime)
379 - CorrelatedFaultID [0..*] (Identifier)
380 - FaultDetails [0..*] (Key-value pair)
382 .. [*] http://www.etsi.org/deliver/etsi_ts/132100_132199/13211102/12.01.00_60/ts_13211102v120100p.pdf
384 * SubscribeFilterClass
386 - VirtualResourceType [0..*] (String)
387 - VirtualResourceID [0..*] (Identifier)
388 - FaultType [0..*] (String)
389 - MinSeverity [0..1] (Integer)
391 * FaultQueryFilterClass: narrows down the FaultQueryRequest, for example it
392 limits the query to certain physical resources, a certain zone, a given fault
393 type/severity/cause, or a specific FaultID.
395 - VirtualResourceType [0..*] (String)
396 - VirtualResourceID [0..*] (Identifier)
397 - FaultType [0..*] (String)
398 - MinSeverity [0..1] (Integer)
399 - EventStartTime [0..1] (Datetime)
400 - EventEndTime [0..1] (Datetime)
402 * PhysicalResourceStateClass:
404 - PhysicalResourceID [1] (Identifier)
405 - PhysicalResourceState [1] (String): mandates the new state of the physical
408 * PhysicalResourceInfoClass:
410 - PhysicalResourceID [1] (Identifier)
411 - PhysicalResourceState [1] (String)
412 - FirmwareVersion [0..1] (String)
413 - HypervisorVersion [0..1] (String)
414 - ZoneID [0..1] (Identifier)
416 * StateQueryFilterClass: narrows down a StateQueryRequest, for example it limits
417 the query to certain physical resources, a certain zone, or a given resource
418 state (e.g., only resources in "maintenance" state).
420 - PhysicalResourceID [1] (Identifier)
421 - PhysicalResourceState [1] (String)
422 - ZoneID [0..1] (Identifier)
426 Detailed northbound interface specification
427 -------------------------------------------
429 This section is specifying the northbound interfaces for fault management and
430 NFVI maintenance between the VIM on the one end and the Consumer and the
431 Administrator on the other ends. For each interface all messages and related
432 information elements are provided.
434 Note: The interface definition will be aligned with current work in ETSI NFV IFA
437 All of the interfaces described below are produced by the VIM and consumed by
438 the Consumer or Administrator.
440 Fault management interface
441 ^^^^^^^^^^^^^^^^^^^^^^^^^^
443 This interface allows the VIM to notify the Consumer about a virtual resource
444 that is affected by a fault, either within the virtual resource itself or by the
445 underlying virtualization infrastructure. The messages on this interface are
446 shown in :numref:`figure13` and explained in detail in the following
449 Note: The information elements used in this section are described in detail in
452 .. figure:: images/figure13.png
456 Fault management NB I/F messages
459 SubscribeRequest (Consumer -> VIM)
460 __________________________________
462 Subscription from Consumer to VIM to be notified about faults of specific
463 resources. The faults to be notified about can be narrowed down using a
468 - SubscribeFilter [1] (SubscribeFilterClass): Optional information to narrow
469 down the faults that shall be notified to the Consumer, for example limit to
470 specific VirtualResourceID(s), severity, or cause of the alarm.
472 SubscribeResponse (VIM -> Consumer)
473 ___________________________________
475 Response to a subscribe request message including information about the
476 subscribed resources, in particular if they are in "fault/error" state.
480 * SubscriptionID [1] (Identifier): Unique identifier for the subscription. It
481 can be used to delete or update the subscription.
482 * VirtualResourceInfo [0..*] (VirtualResourceInfoClass): Provides additional
483 information about the subscribed resources, i.e., a list of the related
484 resources, the current state of the resources, etc.
486 FaultNotification (VIM -> Consumer)
487 ___________________________________
489 Notification about a virtual resource that is affected by a fault, either within
490 the virtual resource itself or by the underlying virtualization infrastructure.
491 After reception of this request, the Consumer will decide on the optimal
492 action to resolve the fault. This includes actions like switching to a hot
493 standby virtual resource, migration of the fault virtual resource to another
494 physical machine, termination of the faulty virtual resource and instantiation
495 of a new virtual resource in order to provide a new hot standby resource.
496 Existing resource management interfaces and messages between the Consumer and
497 the VIM can be used for those actions, and there is no need to define additional
498 actions on the Fault Management Interface.
502 * NotificationID [1] (Identifier): Unique identifier for the notification.
503 * VirtualResourceInfo [1..*] (VirtualResourceInfoClass): List of faulty
504 resources with detailed information about the faults.
506 FaultQueryRequest (Consumer -> VIM)
507 ___________________________________
509 Request to find out about active alarms at the VIM. A FaultQueryFilter can be
510 used to narrow down the alarms returned in the response message.
514 * FaultQueryFilter [1] (FaultQueryFilterClass): narrows down the
515 FaultQueryRequest, for example it limits the query to certain physical
516 resources, a certain zone, a given fault type/severity/cause, or a specific
519 FaultQueryResponse (VIM -> Consumer)
520 ____________________________________
522 List of active alarms at the VIM matching the FaultQueryFilter specified in the
527 * VirtualResourceInfo [0..*] (VirtualResourceInfoClass): List of faulty
528 resources. For each resource all faults including detailed information about
529 the faults are provided.
534 The NFVI maintenance interfaces Consumer-VIM allows the Consumer to subscribe to
535 maintenance notifications provided by the VIM. The related maintenance interface
536 Administrator-VIM allows the Administrator to issue maintenance requests to the
537 VIM, i.e. requesting the VIM to take appropriate actions to empty physical
538 machine(s) in order to execute maintenance operations on them. The interface
539 also allows the Administrator to query the state of physical machines, e.g., in
540 order to get details in the current status of the maintenance operation like a
543 The messages defined in these northbound interfaces are shown in :numref:`figure14`
544 and described in detail in the following subsections.
546 .. figure:: images/figure14.png
550 NFVI maintenance NB I/F messages
552 SubscribeRequest (Consumer -> VIM)
553 __________________________________
555 Subscription from Consumer to VIM to be notified about maintenance operations
556 for specific virtual resources. The resources to be informed about can be
557 narrowed down using a subscribe filter.
561 * SubscribeFilter [1] (SubscribeFilterClass): Information to narrow down the
562 faults that shall be notified to the Consumer, for example limit to specific
563 virtual resource type(s).
565 SubscribeResponse (VIM -> Consumer)
566 ___________________________________
568 Response to a subscribe request message, including information about the
569 subscribed virtual resources, in particular if they are in "maintenance" state.
573 * SubscriptionID [1] (Identifier): Unique identifier for the subscription. It
574 can be used to delete or update the subscription.
575 * VirtualResourceInfo [0..*] (VirtalResourceInfoClass): Provides additional
576 information about the subscribed virtual resource(s), e.g., the ID, type and
577 current state of the resource(s).
579 MaintenanceNotification (VIM -> Consumer)
580 _________________________________________
582 Notification about a physical resource switched to "maintenance" state. After
583 reception of this request, the Consumer will decide on the optimal action to
584 address this request, e.g., to switch to the standby (STBY) configuration.
588 * VirtualResourceInfo [1..*] (VirtualResourceInfoClass): List of virtual
589 resources where the state has been changed to maintenance.
591 StateChangeRequest (Administrator -> VIM)
592 _________________________________________
594 Request to change the state of a list of physical resources, e.g. to
595 "maintenance" state, in order to prepare them for a planned maintenance
600 * PhysicalResourceState [1..*] (PhysicalResourceStateClass)
602 StateChangeResponse (VIM -> Administrator)
603 __________________________________________
605 Response message to inform the Administrator that the requested resources are
606 now in maintenance state (or the operation resulted in an error) and the
607 maintenance operation(s) can be executed.
611 * PhysicalResourceInfo [1..*] (PhysicalResourceInfoClass)
613 StateQueryRequest (Administrator -> VIM)
614 ________________________________________
616 In this procedure, the Administrator would like to get the information about
617 physical machine(s), e.g. their state ("normal", "maintenance"), firmware
618 version, hypervisor version, update status of firmware and hypervisor, etc. It
619 can be used to check the progress during firmware update and the confirmation
620 after update. A filter can be used to narrow down the resources returned in the
625 * StateQueryFilter [1] (StateQueryFilterClass): narrows down the
626 StateQueryRequest, for example it limits the query to certain physical
627 resources, a certain zone, or a given resource state.
629 StateQueryResponse (VIM -> Administrator)
630 _________________________________________
632 List of physical resources matching the filter specified in the
637 * PhysicalResourceInfo [0..*] (PhysicalResourceInfoClass): List of physical
638 resources. For each resource, information about the current state, the
639 firmware version, etc. is provided.
644 This section is listing a first set of blueprints that have been proposed by the
645 Doctor project to the open source community. Further blueprints addressing other
646 gaps identified in Section 4 will be submitted at a later stage of the OPNFV. In
647 this section the following definitions are used:
649 * "Event" is a message emitted by other OpenStack services such as Nova and
650 Neutron and is consumed by the "Notification Agents" in Ceilometer.
651 * "Notification" is a message generated by a "Notification Agent" in Ceilometer
652 based on an "event" and is delivered to the "Collectors" in Ceilometer that
653 store those notifications (as "sample") to the Ceilometer "Databases".
655 Instance State Notification (Ceilometer) [*]_
656 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
658 The Doctor project is planning to handle "events" and "notifications" regarding
659 Resource Status; Instance State, Port State, Host State, etc. Currently,
660 Ceilometer already receives "events" to identify the state of those resources,
661 but it does not handle and store them yet. This is why we also need a new event
662 definition to capture those resource states from "events" created by other
665 This BP proposes to add a new compute notification state to handle events from
666 an instance (server) from nova. It also creates a new meter "instance.state" in
669 .. [*] https://etherpad.opnfv.org/p/doctor_bps
671 Event Publisher for Alarm (Ceilometer) [*]_
672 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
674 **Problem statement:**
676 The existing "Alarm Evaluator" in OpenStack Ceilometer is periodically
677 querying/polling the databases in order to check all alarms independently from
678 other processes. This is adding additional delay to the fault notification
679 send to the Consumer, whereas one requirement of Doctor is to react on faults
682 The existing message flow is shown in :numref:`figure12`: after receiving
683 an "event", a "notification agent" (i.e. "event publisher") will send a
684 "notification" to a "Collector". The "collector" is collecting the
685 notifications and is updating the Ceilometer "Meter" database that is storing
686 information about the "sample" which is capured from original "event". The
687 "Alarm Evaluator" is periodically polling this databases then querying "Meter"
688 database based on each alarm configuration.
690 In the current Ceilometer implementation, there is no possibility to directly
691 trigger the "Alarm Evaluator" when a new "event" was received, but the "Alarm
692 Evaluator" will only find out that requires firing new notification to the
693 Consumer when polling the database.
695 **Change/feature request:**
697 This BP proposes to add a new "event publisher for alarm", which is bypassing
698 several steps in Ceilometer in order to avoid the polling-based approach of
699 the existing Alarm Evaluator that makes notification slow to users.
701 After receiving an "(alarm) event" by listening on the Ceilometer message
702 queue ("notification bus"), the new "event publisher for alarm" immediately
703 hands a "notification" about this event to a new Ceilometer component
704 "Notification-driven alarm evaluator" proposed in the other BP (see Section
707 Note, the term "publisher" refers to an entity in the Ceilometer architecture
708 (it is a "notification agent"). It offers the capability to provide
709 notifications to other services outside of Ceilometer, but it is also used to
710 deliver notifications to other Ceilometer components (e.g. the "Collectors")
711 via the Ceilometer "notification bus".
713 **Implementation detail**
715 * "Event publisher for alarm" is part of Ceilometer
716 * The standard AMQP message queue is used with a new topic string.
717 * No new interfaces have to be added to Ceilometer.
718 * "Event publisher for Alarm" can be configured by the Administrator of
719 Ceilometer to be used as "Notification Agent" in addition to the existing
721 * Existing alarm mechanisms of Ceilometer can be used allowing users to
722 configure how to distribute the "notifications" transformed from "events",
723 e.g. there is an option whether an ongoing alarm is re-issued or not
726 .. [*] https://etherpad.opnfv.org/p/doctor_bps
728 Notification-driven alarm evaluator (Ceilometer) [*]_
729 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
731 **Problem statement:**
733 The existing "Alarm Evaluator" in OpenStack Ceilometer is periodically
734 querying/polling the databases in order to check all alarms independently from
735 other processes. This is adding additional delay to the fault notification send
736 to the Consumer, whereas one requirement of Doctor is to react on faults as fast
739 **Change/feature request:**
741 This BP is proposing to add an alternative "Notification-driven Alarm Evaluator"
742 for Ceilometer that is receiving "notifications" sent by the "Event Publisher
743 for Alarm" described in the other BP. Once this new "Notification-driven Alarm
744 Evaluator" received "notification", it finds the "alarm" configurations which
745 may relate to the "notification" by querying the "alarm" database with some keys
746 i.e. resource ID, then it will evaluate each alarm with the information in that
749 After the alarm evaluation, it will perform the same way as the existing "alarm
750 evaluator" does for firing alarm notification to the Consumer. Similar to the
751 existing Alarm Evaluator, this new "Notification-driven Alarm Evaluator" is
752 aggregating and correlating different alarms which are then provided northbound
753 to the Consumer via the OpenStack "Alarm Notifier". The user/administrator can
754 register the alarm configuration via existing Ceilometer API [*]_. Thereby, he
755 can configure whether to set an alarm or not and where to send the alarms to.
757 **Implementation detail**
759 * The new "Notification-driven Alarm Evaluator" is part of Ceilometer.
760 * Most of the existing source code of the "Alarm Evaluator" can be re-used to
762 * No additional application logic is needed
763 * It will access the Ceilometer Databases just like the existing "Alarm
765 * Only the polling-based approach will be replaced by a listener for
766 "notifications" provided by the "Event Publisher for Alarm" on the Ceilometer
768 * No new interfaces have to be added to Ceilometer.
771 .. [*] https://etherpad.opnfv.org/p/doctor_bps
772 .. [*] https://wiki.openstack.org/wiki/Ceilometer/Alerting
774 Report host fault to update server state immediately (Nova) [*]_
775 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
777 **Problem statement:**
779 * Nova state change for failed or unreachable host is slow and does not reliably
780 state host is down or not. This might cause same server instance to run twice
781 if action taken to evacuate instance to another host.
782 * Nova state for server(s) on failed host will not change, but remains active
783 and running. This gives the user false information about server state.
784 * VIM northbound interface notification of host faults towards VNFM and NFVO
785 should be in line with OpenStack state. This fault notification is a Telco
786 requirement defined in ETSI and will be implemented by OPNFV Doctor project.
787 * Openstack user cannot make HA actions fast and reliably by trusting server
788 state and host state.
792 There needs to be a new API for Admin to state host is down. This API is used to
793 mark services running in host down to reflect the real situation.
795 Example on compute node is:
797 * When compute node is up and running:::
799 vm_state: activeand power_state: running
800 nova-compute state: up status: enabled
802 * When compute node goes down and new API is called to state host is down:::
804 vm_state: stopped power_state: shutdown
805 nova-compute state: down status: enabled
809 There is no attractive alternative to detect all different host faults than to
810 have an external tool to detect different host faults. For this kind of tool to
811 exist there needs to be new API in Nova to report fault. Currently there must be
812 some kind of workarounds implemented as cannot trust or get the states from
813 OpenStack fast enough.
815 .. [*] https://blueprints.launchpad.net/nova/+spec/update-server-state-immediately
820 This section lists some BPs related to Doctor, but proposed by drafters outside
823 pacemaker-servicegroup-driver [*]_
824 __________________________________
826 This BP will detect and report host down quite fast to OpenStack. This however
827 might not work properly for example when management network has some problem and
828 host reported faulty while VM still running there. This might lead to launching
829 same VM instance twice causing problems. Also NB IF message needs fault reason
830 and for that the source needs to be a tool that detects different kind of faults
831 as Doctor will be doing. Also this BP might need enhancement to change server
832 and service states correctly.
834 .. [*] https://blueprints.launchpad.net/nova/+spec/pacemaker-servicegroup-driver
837 vim: set tabstop=4 expandtab textwidth=80: