Difference between revisions of "Graffiti"
Travis Tripp (talk | contribs) m (→Initial use case) |
Travis Tripp (talk | contribs) (→Current Status) |
||
(88 intermediate revisions by the same user not shown) | |||
Line 1: | Line 1: | ||
+ | |||
+ | == What's in my cloud? == | ||
+ | |||
+ | I've got a lot of resources in my cloud. | ||
+ | |||
+ | * How do I find what I need? | ||
+ | * How do I describe what I have? | ||
+ | |||
+ | At its most basic concept, Graffiti's intent is to enable better metadata collaboration across services and projects so that OpenStack users can take advantage of an Enhanced Platform Awareness. | ||
+ | |||
+ | == Current Status == | ||
+ | |||
+ | The Graffiti project was proposed at the Atlanta (Juno) OpenStack summit. Since then the concepts have been adopted and implemented as part of multiple different OpenStack Projects. | ||
+ | |||
+ | * Glance Metadata Definition Catalog | ||
+ | ** http://docs.openstack.org/developer/glance/metadefs-concepts.html | ||
+ | ** https://github.com/openstack/glance/tree/master/etc/metadefs | ||
+ | ** https://youtu.be/zJpHXdBOoeM | ||
+ | * Searchlight | ||
+ | ** http://launchpad.net/searchlight | ||
+ | ** https://wiki.openstack.org/wiki/Searchlight | ||
+ | * Nova features | ||
+ | ** Scheduling filters like Numa topology | ||
+ | * Horizon features | ||
+ | ** An admin UI for managing the catalog | ||
+ | *** (Admin —> Metadata Definitions) (Kilo) | ||
+ | ** A widget for associating metadata to different resources | ||
+ | *** (Update Metadata action on each row item below) | ||
+ | *** admin -> images (Juno) | ||
+ | *** admin -> flavors (Kilo) | ||
+ | *** admin —> Host Aggregates (Kilo) | ||
+ | *** project —> images (Liberty) | ||
+ | *** project —> instances (Mitaka) | ||
+ | ** The ability to add metadata at launch time | ||
+ | *** project —> Launch Instance (ng launch instance enabled) (Mitaka) | ||
+ | |||
+ | The following information provides much of the background information on where these concepts originated. | ||
== Overview == | == Overview == | ||
− | |||
− | We | + | A challenge we've experienced with using OpenStack is discovering, sharing, and correlating metadata across services and different types of resources. We believe this affects both end users and administrators. |
+ | |||
+ | For end users, we feel like doing basic tasks like launching instances is too technical for end users and require too much pre-existing knowledge of OpenStack concepts. For example, you should be able to just specify categories like "Big Data" or an "OS Family" and then let the system find the boot source for you, whether that is an image, snapshot, or volume. It should also allow finer grained filtering like filtering on specific versions of software that you want. | ||
+ | |||
+ | For administrators, we’d like there to be an easier way to meaningfully collaborate on properties across host aggregates, flavors, images, volumes, or other cloud resources. | ||
+ | |||
+ | Various OpenStack services provide techniques to abstract low level resource selection to one level higher, such as flavors, volume types, or artifact types. These resource abstractions often allow "metadata" in terms of key-value pair properties to further specialize and describe instances of each resource type. However, collaborating on those properties can be a disconnected and difficult process. This often involves searching wikis and opening the source code. In addition, the metadata properties often need to be correlated across several different services. It becomes more difficult as a cloud's scale grows and the number of resources being managed increases. | ||
+ | |||
+ | We, HP and Intel, believe that both of the above problems come back to needing a better way for users to collaborate on metadata across services and resource types. We started a project called Graffiti to explore ideas and concepts for how to make this easier and more approachable for end users. Please join with us to help move forward together as a community! | ||
+ | |||
+ | We believe that we can make some immediate improvements in Horizon, but that they can't be achieved through Horizon alone and that the benefits should extend to the API and CLI interactions as well. Better cross service collaboration and consistency on metadata should provide benefits that can be leveraged by other projects such as scheduling, reservation, orchestration, and policy enforcement. | ||
+ | |||
+ | === Terminology Note === | ||
− | We | + | We think the term "metadata" is a somewhat unapproachable term, so we have been exploring with the concept of a "capability". A capability can simply be thought of as a named "tag" which may or may not have properties. The idea is that a user can simply "tag" a capability onto various cloud resources such as images, volumes, host aggregates, flavors, and so on. To the end user, the exact mechanism for how the data is stored is handled for them. |
− | + | === Screencasts === | |
− | + | To help explain the ideas of the project, we have a quick screencast demonstrating the concepts running under POC code. Please take a look! | |
− | + | * [https://www.youtube.com/watch?v=f0SZtPgcxk4| Concept Overview] | |
− | + | * [https://youtu.be/zJpHXdBOoeM Availability as of the mitaka release in Horizon and Glance] | |
− | |||
− | |||
+ | === Usage Concepts === | ||
− | + | # Load your metadata definitions (sometimes called properties, tags, or capabilities) | |
+ | ## Into the central metadata catalog | ||
+ | # Update the resources in the cloud with your tags and capabilities | ||
+ | # Let users find the resources with your desired tags and capabilities | ||
− | == | + | == Design Concepts == |
− | + | Additional architecture concepts on the [[Graffiti/Architecture|Architecture]] page. | |
− | == | + | === Juno Summit Design Sessioɲ === |
− | = | + | POC Demo reviewː |
+ | ̽ * https://www.youtube.com/watch?v=Dhrthnq1bnw | ||
+ | http://sched.co/1m7wghx | ||
+ | * Etherpadː https://etherpad.openstack.org/p/juno-summit-graffiti | ||
− | === | + | === IRC === |
− | + | The various features are maintained by teams in the following IRC channels on [http://freenode.net/ Freenode]. | |
− | + | #openstack-searchlight | |
+ | #openstack-horizon | ||
+ | #openstack-glance | ||
− | + | === Development === | |
− | * | + | * Open source under Apache 2.0 |
− | * [https://github.com/stackforge/graffiti | + | * [https://github.com/stackforge/graffiti Graffiti POC API Service Source Repository] - No Longer Maintained (See Glance, Horizon, Searchlight) |
Latest revision as of 20:26, 8 January 2021
Contents
What's in my cloud?
I've got a lot of resources in my cloud.
- How do I find what I need?
- How do I describe what I have?
At its most basic concept, Graffiti's intent is to enable better metadata collaboration across services and projects so that OpenStack users can take advantage of an Enhanced Platform Awareness.
Current Status
The Graffiti project was proposed at the Atlanta (Juno) OpenStack summit. Since then the concepts have been adopted and implemented as part of multiple different OpenStack Projects.
- Glance Metadata Definition Catalog
- Searchlight
- Nova features
- Scheduling filters like Numa topology
- Horizon features
- An admin UI for managing the catalog
- (Admin —> Metadata Definitions) (Kilo)
- A widget for associating metadata to different resources
- (Update Metadata action on each row item below)
- admin -> images (Juno)
- admin -> flavors (Kilo)
- admin —> Host Aggregates (Kilo)
- project —> images (Liberty)
- project —> instances (Mitaka)
- The ability to add metadata at launch time
- project —> Launch Instance (ng launch instance enabled) (Mitaka)
- An admin UI for managing the catalog
The following information provides much of the background information on where these concepts originated.
Overview
A challenge we've experienced with using OpenStack is discovering, sharing, and correlating metadata across services and different types of resources. We believe this affects both end users and administrators.
For end users, we feel like doing basic tasks like launching instances is too technical for end users and require too much pre-existing knowledge of OpenStack concepts. For example, you should be able to just specify categories like "Big Data" or an "OS Family" and then let the system find the boot source for you, whether that is an image, snapshot, or volume. It should also allow finer grained filtering like filtering on specific versions of software that you want.
For administrators, we’d like there to be an easier way to meaningfully collaborate on properties across host aggregates, flavors, images, volumes, or other cloud resources.
Various OpenStack services provide techniques to abstract low level resource selection to one level higher, such as flavors, volume types, or artifact types. These resource abstractions often allow "metadata" in terms of key-value pair properties to further specialize and describe instances of each resource type. However, collaborating on those properties can be a disconnected and difficult process. This often involves searching wikis and opening the source code. In addition, the metadata properties often need to be correlated across several different services. It becomes more difficult as a cloud's scale grows and the number of resources being managed increases.
We, HP and Intel, believe that both of the above problems come back to needing a better way for users to collaborate on metadata across services and resource types. We started a project called Graffiti to explore ideas and concepts for how to make this easier and more approachable for end users. Please join with us to help move forward together as a community!
We believe that we can make some immediate improvements in Horizon, but that they can't be achieved through Horizon alone and that the benefits should extend to the API and CLI interactions as well. Better cross service collaboration and consistency on metadata should provide benefits that can be leveraged by other projects such as scheduling, reservation, orchestration, and policy enforcement.
Terminology Note
We think the term "metadata" is a somewhat unapproachable term, so we have been exploring with the concept of a "capability". A capability can simply be thought of as a named "tag" which may or may not have properties. The idea is that a user can simply "tag" a capability onto various cloud resources such as images, volumes, host aggregates, flavors, and so on. To the end user, the exact mechanism for how the data is stored is handled for them.
Screencasts
To help explain the ideas of the project, we have a quick screencast demonstrating the concepts running under POC code. Please take a look!
Usage Concepts
- Load your metadata definitions (sometimes called properties, tags, or capabilities)
- Into the central metadata catalog
- Update the resources in the cloud with your tags and capabilities
- Let users find the resources with your desired tags and capabilities
Design Concepts
Additional architecture concepts on the Architecture page.
Juno Summit Design Sessioɲ
POC Demo reviewː ̽ * https://www.youtube.com/watch?v=Dhrthnq1bnw
IRC
The various features are maintained by teams in the following IRC channels on Freenode.
#openstack-searchlight #openstack-horizon #openstack-glance
Development
- Open source under Apache 2.0
- Graffiti POC API Service Source Repository - No Longer Maintained (See Glance, Horizon, Searchlight)