Jump to: navigation, search

Difference between revisions of "PersonalHomepage"

Line 22: Line 22:
 
== User stories ==
 
== User stories ==
  
−
# As a systems integrator, I need to retrieve usage data so that I can properly bill my customers.
+
# As a systems integrator, I need to retrieve usage data so that I can properly bill my customers.
−
# As a systems integrator or enterprise cloud administrator, I want to monitor usage data over time so that I can optimize the utilization of my resources.
+
# As a systems integrator or enterprise cloud administrator, I want to monitor usage data over time so that I can optimize the utilization of my resources.
−
# As a systems integrator, I need to provide usage data to multiple third-party systems without building custom interfaces to each one so that I can efficiently utilize my manpower.
+
# As a systems integrator, I need to provide usage data to multiple third-party systems without building custom interfaces to each one so that I can efficiently utilize my manpower.
  
 
== Design ==
 
== Design ==
Line 33: Line 33:
  
 
* Instances
 
* Instances
−
* * Memory increase/decrease
+
** Memory increase/decrease
−
* * Disk increase/decrease
+
** Disk increase/decrease
−
* * Image mounted
+
** Image mounted
−
* * BW consumed
+
** BW consumed
 
* Volumes
 
* Volumes
−
* * Created/deleted
+
** Created/deleted
−
* * Attached/detached
+
** Attached/detached
 
* Images:
 
* Images:
−
* * Created/Deleted
+
** Created/Deleted
 
* Floating IPs
 
* Floating IPs
−
* * Associated/dissasociated (to an instance)
+
** Associated/dissasociated (to an instance)
−
* * Allocated/released (by a user/tenant)
+
** Allocated/released (by a user/tenant)
 
* Snapshots
 
* Snapshots
 
* Fixed Networks
 
* Fixed Networks
Line 57: Line 57:
  
 
{| border="1" cellpadding="2" cellspacing="0"
 
{| border="1" cellpadding="2" cellspacing="0"
−
|  Type  
+
|  '''Type'''
−
|  Resource  
+
|  '''Resource'''
 +
|-
 +
|  compute.instance.create.start/.end **
 +
|  Instance
 +
|-
 +
|  compute.instance.delete.start/.end **
 +
|  Instance
 +
|-
 +
|  compute.instance.rebuild.start/.end **
 +
|  Instance
 +
|-
 +
|  compute.instance.resize.prep.start/.end **
 +
|  Instance
 +
|-
 +
|  compute.instance.resize.confirm.start/.end **
 +
|  Instance
 +
|-
 +
|  compute.instance.resize.revert.start/.end **
 +
|  Instance
 +
|-
 +
|  compute.instance.exists **
 +
|  Instance Audit
 +
|-
 +
|  compute.instance.reboot.start/.end ***
 +
|  Instance
 +
|-
 +
|  compute.volume.create.start/.end
 +
|  Volume
 +
|-
 +
|  compute.volume.delete.start/.end
 +
|  Volume
 +
|-
 +
|  compute.volume.attached.start/.end
 +
|  Volume
 +
|-
 +
|  compute.volume.detached.start/.end
 +
|  Volume
 +
|-
 +
|  compute.volume.exists
 +
|  Volume Audit
 +
|-
 +
|  compute.image.create.start/.end
 +
|  Image
 +
|-
 +
|  compute.image.delete.start/.end
 +
|  Image
 +
|-
 +
|  compute.image.exists
 +
|  Image Audit
 +
|-
 +
|  compute.ip.allocate.start/.end
 +
|  IP
 +
|-
 +
|  compute.ip.release.start/.end
 +
|  IP
 +
|-
 +
|  compute.instance.create_ip ***
 +
|  IP
 +
|-
 +
|  compute.instance.delete_ip ***
 +
|  IP
 +
|-
 +
|  compute.ip.exists
 +
|  IP Audit
 
|}
 
|}
−
| compute.instance.create.start/.end \*\* | Instance | Usage notification upon creation of a new instance |
 
−
| compute.instance.delete.start/.end *\* | Instance | Usage notification upon deletion of an instance |
 
−
| compute.instance.rebuild.start/.end *\* | Instance | Usage notification upon rebuild of instance |
 
−
| compute.instance.resize.prep.start/.end *\* | Instance | Usage notification when a resize/migrate of an instance is begun |
 
−
| compute.instance.resize.confirm.start/.end *\* | Instance | Usage notification when a resize/migration of an instance has been confirmed by the user |
 
−
| compute.instance.resize.revert.start/.end *\* | Instance | Usage notification when a resize/migration of an instance has been reverted by the user |
 
−
| compute.instance.exists *\* | Instance Audit | Periodic usage notification generated by the instance-usage-audit cron job. |
 
−
| {color:#ff0000}compute.instance.reboot.start/.end *{color} | Instance | Usage notification when a reboot of the instance has been executed |
 
−
| compute.volume.create.start/.end | Volume | Creation or a new volume notification |
 
−
| compute.volume.delete.start/.end | Volume | Deletion of a volume notification |
 
−
| compute.volume.attached.start/.end | Volume | Usage notification upon attaching of a volume to an instance |
 
−
| compute.volume.detached.start/.end | Volume | Usage notification upon a volume has been detached from an instance |
 
−
| compute.volume.exists | Volume Audit | Periodic usage notification generated by the volume-usage-audit cron job. |
 
−
| compute.image.create.start/.end | Image | Creation of a new image |
 
−
| compute.image.delete.start/.end | Image | Deletion of an image |
 
−
| compute.image.exists | Image Audit | Periodic usage notification generated by the image-usage-audit cron job. |
 
−
| compute.ip.allocate.start/.end | IP | Allocation of an IP address notification |
 
−
| compute.ip.release.start/.end | IP | Deletion of an IP address notification |
 
−
| compute.ip.associate.start/.end {color:#ff0000}(compute.instance.create_ip)\*{color} | IP | Association of an IP address to an instance |
 
−
| compute.ip.disassociate.start/.end {color:#ff0000}(compute.instance.delete_ip)\*{color} | IP | Disassociation of an IP address to an instance |
 
−
| compute.ip.exists | IP Audit | Periodic usage notification generated by the ip-usage-audit cron job. |
 
  
−
\*\* Events from the previous version: [http://wiki.openstack.org/SystemUsageData]
+
* * Events from the previous version: [http://wiki.openstack.org/SystemUsageData]
 +
* ** Seen in the reviewed code [nova/compute/manager.py|https://review.openstack.org/#patch,sidebyside,2460,10,nova/compute/manager.py]
  
−
{color:#ff0000}\* Seen in the reviewed code {color}[nova/compute/manager.py|https://review.openstack.org/#patch,sidebyside,2460,10,nova/compute/manager.py]
+
Example of a reboot event:
  
−
Example of a reboot event:
+
<pre><nowiki>#!highlight json
−
{code}
 
 
{
 
{
 
   "_context_roles": ["admin"],
 
   "_context_roles": ["admin"],
Line 120: Line 162:
 
   "_context_remote_address": null
 
   "_context_remote_address": null
 
}
 
}
−
{code}
+
</nowiki></pre>
 +
 
  
 
Create instance event:
 
Create instance event:
−
{code}
+
 
 +
<pre><nowiki>#!highlight json
 
{
 
{
 
   "_context_roles": ["admin"],
 
   "_context_roles": ["admin"],
Line 163: Line 207:
 
   "_context_remote_address": null
 
   "_context_remote_address": null
 
}
 
}
−
{code}
+
</nowiki></pre>
 +
 
  
 
=== Resources Model ===
 
=== Resources Model ===

Revision as of 08:48, 30 March 2012

  • Launchpad Entry: NovaSpec:nova-notifications
  • Created: 28 Sep 2011
  • Contributors: Juan Lebrijo

Summary

Nova administrators want data on system usage for billing, chargeback, or monitoring purposes.

Release Note

This paper is based on "System Usage Data" http://wiki.openstack.org/SystemUsageData implemented for Essex realease.

We are planning to implement this features for Folsom

Rationale

The best way to gather this data is from a notifications system based on AMQP queues (RabbitMQ messaging System).

Then the main challenge of this BluePrint is to define the most useful events to be transmitted from OpenStack subsystems to AMQP queues. In order to be consumed by billing, chargeback or monitoring systems.

User stories

  1. As a systems integrator, I need to retrieve usage data so that I can properly bill my customers.
  2. As a systems integrator or enterprise cloud administrator, I want to monitor usage data over time so that I can optimize the utilization of my resources.
  3. As a systems integrator, I need to provide usage data to multiple third-party systems without building custom interfaces to each one so that I can efficiently utilize my manpower.

Design

There are two events generated for every activity <activity>.start and <activity>.end. The "start" one specifies the beginning of the activity, and the "end" one specifies the success of the activity.

The Events resources are the objects in the platform that we are interested to scan:

  • Instances
    • Memory increase/decrease
    • Disk increase/decrease
    • Image mounted
    • BW consumed
  • Volumes
    • Created/deleted
    • Attached/detached
  • Images:
    • Created/Deleted
  • Floating IPs
    • Associated/dissasociated (to an instance)
    • Allocated/released (by a user/tenant)
  • Snapshots
  • Fixed Networks
  • Tenants
  • Users
  • Inbound traffic
  • Outbound traffic

Implementation

Event Types

Type Resource
compute.instance.create.start/.end ** Instance
compute.instance.delete.start/.end ** Instance
compute.instance.rebuild.start/.end ** Instance
compute.instance.resize.prep.start/.end ** Instance
compute.instance.resize.confirm.start/.end ** Instance
compute.instance.resize.revert.start/.end ** Instance
compute.instance.exists ** Instance Audit
compute.instance.reboot.start/.end *** Instance
compute.volume.create.start/.end Volume
compute.volume.delete.start/.end Volume
compute.volume.attached.start/.end Volume
compute.volume.detached.start/.end Volume
compute.volume.exists Volume Audit
compute.image.create.start/.end Image
compute.image.delete.start/.end Image
compute.image.exists Image Audit
compute.ip.allocate.start/.end IP
compute.ip.release.start/.end IP
compute.instance.create_ip *** IP
compute.instance.delete_ip *** IP
compute.ip.exists IP Audit

Example of a reboot event: 

#!highlight json
{
  "_context_roles": ["admin"],
  "_context_request_id": "req-b41fca68-5fe4-4c9a-b3ba-956b72bae971",
  "event_type": "compute.instance.reboot.end",
  "timestamp": "2012-02-29 00:07:23.442571",
  "_context_auth_token": null,
  "_context_user_id": null,
  "payload": {
    "state_description": "rebooting_hard",
    "display_name": "instance03",
    "disk_gb": 0,
    "tenant_id": "54deadcc2c724dadbb3f59b139c97c09",
    "created_at": "2012-02-28 23:56:10",
    "instance_id": "8c12bedb-1817-4543-9223-2484b138dc66",
    "instance_type_id": 2,
    "memory_mb": 512,
    "instance_type": "m1.tiny",
    "state": "active",
    "user_id": "admin",
    "launched_at": "2012-02-28 23:56:17",
    "image_ref_url": "http:\/\/192.168.10.179:9292\/images\/9c5bbe68-04eb-4687-ae90-f96b876c9d50"
  },
  "_context_read_deleted": "no",
  "priority": "INFO",
  "_context_strategy": "noauth",
  "_context_is_admin": true,
  "_context_project_id": null,
  "_context_timestamp": "2012-02-29T00:07:23.442607",
  "publisher_id": "compute.devstack",
  "message_id": "88d14623-fa5a-4e2e-aff6-3289b7ae72dd",
  "_context_remote_address": null
}


Create instance event: 

#!highlight json
{
  "_context_roles": ["admin"],
  "_context_request_id": "req-67c9cc6a-c6a0-4bd6-901d-4cab5fac2704",
  "event_type": "compute.instance.create.end",
  "timestamp": "2012-02-29 00:16:51.844539",
  "_context_auth_token": null,
  "_context_user_id": null,
  "payload": {
    "state_description": "",
    "display_name": "instance07",
    "disk_gb": 0,
    "tenant_id": "464e0e5fac1747e1b81f482d80525ea6",
    "created_at": "2012-02-29 00:16:43",
    "instance_id": "576b3a16-b8ee-4105-b84c-623fbbda1f04",
    "instance_type_id": 2,
    "memory_mb": 512,
    "instance_type": "m1.tiny",
    "state": "active",
    "user_id": "admin",
    "fixed_ips": [{
      "meta": {},
      "type": "fixed",
      "version": 4,
      "address": "10.0.0.3",
      "floating_ips": []
    }],
    "launched_at": "2012-02-29 00:16:51.564723",
    "image_ref_url": "http:\/\/192.168.10.179:9292\/images\/9c5bbe68-04eb-4687-ae90-f96b876c9d50"
  },
  "_context_read_deleted": "no",
  "priority": "INFO",
  "_context_strategy": "noauth",
  "_context_is_admin": true,
  "_context_project_id": null,
  "_context_timestamp": "2012-02-29T00:16:51.844580",
  "publisher_id": "compute.devstack",
  "message_id": "be0d96ba-73ce-4a37-b6a9-923eadd8f86a",
  "_context_remote_address": null
}


Resources Model

Instance

Instance Audit

All instance attributes and:

Image

Volume

IP

NovaSpec

Retrieved from "https://wiki.openstack.org/w/index.php?title=PersonalHomepage&oldid=12655"