@fluidframework/container-runtime
Data Virtualization For DataStores (Work in Progress)
It's a capability to exclude some content from initial snapshot (used when loading container) and thus improve boot
performance of a container. Excluded content could be loaded at a later time when it's required.
Motivation for Data Virtualization
This section talks about how the system used to work before Data virtualization. Currently, the content of whole file
is downloaded in one go. Due to limitation of data virtualization, FF holds all blobs in snapshot as those might be
required in the future. Any delayed loading (through FF APIs) results in loading state of datastores at a sequence
number of snapshot we booted from, up until the current sequence number by applying the pending ops for that datastore.
While application may choose not to load some data stores immediately on boot (and realize some saving in time and
memory by not allocating appropriate app state for such datastores), FF still pays the costs for such content. It also
continues to pay the cost for all such content indefinitely, even if those datastores were loaded.
Improvement with Data Virtualization
With this, we will provide a capability to:
- Exclude some sub-trees from snapshot payload, thus allowing faster transfer times / boot times and smaller initial
memory footprint.
- Ability to delay-load data stores later.
Container Runtime Apis like IContainerRuntimeBase.createDataStore and IContainerRuntimeBase.createDetachedDataStore
provides an argument loadingGroupId
which allows apps to mark a datastore at time of creation currently. Every data
store is assigned a groupID. Not providing groupID (on API) means that default ID is used. This groupId represents the
group of the datastore within a container or its snapshot. When a container is loaded initially, only datastores which
belongs to default group are fetched from service and can be loaded on demand when requested by user. This decreases
the amount of data which needs to be fetched during load and hence provides faster boot times for the container.
Attempting to load any datastore within a non-default group results in fetching all content/datastores marked with same
groupId. So, one network will be required to fetch content for a group when a datastore from a group is requested by an
application.
In advanced app scenarios, app would want to make datastores with a specific group Id, based on how it wants to load a
certain group at once, and not load the datastores that aren't part of the group. By effectively using groupID, app
will be able to improve boot times by not fetching unnecessary groups at load.
So to summarize, when datastore is assigned to a group, content of such data store will not be loaded with initial load
of container. It will be loaded only when any datastore with such groupID is realized.
This will improve the boot perf. Data virtualization or providing the loadingGroupId
will however decrease the
performance of loading of those datastores as one network call would be required before loading. However,
providing same loadingGroupId
to put some data stores in same group, would improve performance for their loading as
compared to providing a different group Id to each of these datastores as then one network call will be required to
fetch snapshot for that group of datastores rather than one network call for each datastore. So, the datastores which
can get fairly big in size content wise and which are not required to be loaded on boot, can be put under a non-default
groupId.
Using Fluid Framework libraries
When taking a dependency on a Fluid Framework library, we recommend using a ^
(caret) version range, such as ^1.3.4
.
While Fluid Framework libraries may use different ranges with interdependencies between other Fluid Framework libraries,
library consumers should always prefer ^
.
Installation
To get started, install the package by running the following command:
npm i @fluidframework/container-runtime
API Documentation
API documentation for @fluidframework/container-runtime is available at https://fluidframework.com/docs/apis/container-runtime.
Contribution Guidelines
There are many ways to contribute to Fluid.
Detailed instructions for working in the repo can be found in the Wiki.
This project has adopted the Microsoft Open Source Code of Conduct.
For more information see the Code of Conduct FAQ or contact opencode@microsoft.com with any additional questions or comments.
This project may contain Microsoft trademarks or logos for Microsoft projects, products, or services.
Use of these trademarks or logos must follow Microsoft’s Trademark & Brand Guidelines.
Use of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship.
Help
Not finding what you're looking for in this README? Check out our GitHub
Wiki or fluidframework.com.
Still not finding what you're looking for? Please file an
issue.
Thank you!
Trademark
This project may contain Microsoft trademarks or logos for Microsoft projects, products, or services.
Use of these trademarks or logos must follow Microsoft's Trademark & Brand Guidelines.
Use of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship.