feat: New Fabric sample datasets (#2219)

* add two new fabric sample datasets.

* Update Fabric Home with two sample datasets. One regular and one for vector search.

* Update specs for sample data container

* Add telemetry instead of console log

* Add sampleDataFile to telemetry when importing sample data

---------

Co-authored-by: Mark Brown <mjbrown@microsoft.com>
Co-authored-by: Laurent Nguyen <languye@microsoft.com>
This commit is contained in:
Laurent Nguyen
2025-10-03 17:31:05 +02:00
committed by GitHub
parent 909957a9a1
commit cd27814fad
10 changed files with 388404 additions and 26 deletions

View File

@@ -42,7 +42,7 @@ import {
} from "Explorer/Panes/AddCollectionPanel/AddCollectionPanelUtility";
import { useSidePanel } from "hooks/useSidePanel";
import { useTeachingBubble } from "hooks/useTeachingBubble";
import { isFabricNative } from "Platform/Fabric/FabricUtil";
import { DEFAULT_FABRIC_NATIVE_CONTAINER_THROUGHPUT, isFabricNative } from "Platform/Fabric/FabricUtil";
import React from "react";
import { CollectionCreation } from "Shared/Constants";
import { Action } from "Shared/Telemetry/TelemetryConstants";
@@ -52,7 +52,6 @@ import { getCollectionName } from "Utils/APITypeUtils";
import { isCapabilityEnabled, isServerlessAccount, isVectorSearchEnabled } from "Utils/CapabilityUtils";
import { getUpsellMessage } from "Utils/PricingUtils";
import { ValidCosmosDbIdDescription, ValidCosmosDbIdInputPattern } from "Utils/ValidationUtils";
import * as AutoPilotUtils from "../../../Utils/AutoPilotUtils";
import { CollapsibleSectionComponent } from "../../Controls/CollapsiblePanel/CollapsibleSectionComponent";
import { ThroughputInput } from "../../Controls/ThroughputInput/ThroughputInput";
import { ContainerSampleGenerator } from "../../DataSamples/ContainerSampleGenerator";
@@ -1355,8 +1354,7 @@ export class AddCollectionPanel extends React.Component<AddCollectionPanelProps,
// Throughput
if (isFabricNative()) {
// Fabric Native accounts are always autoscale and have a fixed throughput of 5K
autoPilotMaxThroughput = AutoPilotUtils.autoPilotThroughput5K;
autoPilotMaxThroughput = DEFAULT_FABRIC_NATIVE_CONTAINER_THROUGHPUT;
offerThroughput = undefined;
} else if (databaseLevelThroughput) {
if (this.state.createNewDatabase) {

View File

@@ -3,12 +3,15 @@
*/
import { Link, makeStyles, tokens } from "@fluentui/react-components";
import { DocumentAddRegular, LinkMultipleRegular, OpenRegular } from "@fluentui/react-icons";
import { SampleDataImportDialog } from "Explorer/SplashScreen/SampleDataImportDialog";
import { SampleDataConfiguration, SampleDataImportDialog } from "Explorer/SplashScreen/SampleDataImportDialog";
import { SampleDataFile } from "Explorer/SplashScreen/SampleUtil";
import { CosmosFluentProvider } from "Explorer/Theme/ThemeUtil";
import { isFabricNative, isFabricNativeReadOnly } from "Platform/Fabric/FabricUtil";
import * as React from "react";
import { userContext } from "UserContext";
import AzureOpenAiIcon from "../../../images/AzureOpenAi.svg";
import CosmosDbBlackIcon from "../../../images/CosmosDB_black.svg";
import GithubIcon from "../../../images/github-black-and-white.svg";
import Explorer from "../Explorer";
export interface SplashScreenProps {
@@ -26,11 +29,11 @@ const useStyles = makeStyles({
fontWeight: "bold",
},
buttonsContainer: {
width: "584px",
width: "760px",
margin: "auto",
display: "grid",
padding: "16px",
gridTemplateColumns: "repeat(3, 1fr)",
gridTemplateColumns: "repeat(4, 1fr)",
gap: "10px",
gridAutoRows: "minmax(184px, auto)",
},
@@ -53,6 +56,15 @@ const useStyles = makeStyles({
},
},
three: {
gridColumn: "4",
gridRow: "1",
"& img": {
width: "32px",
height: "32px",
margin: "auto",
},
},
four: {
gridColumn: "3",
gridRow: "2",
"& svg": {
@@ -61,6 +73,15 @@ const useStyles = makeStyles({
margin: "auto",
},
},
five: {
gridColumn: "4",
gridRow: "2",
"& img": {
width: "32px",
height: "32px",
margin: "auto",
},
},
single: {
gridColumn: "1 / 4",
gridRow: "1 / 3",
@@ -132,6 +153,8 @@ const FabricHomeScreenButton: React.FC<FabricHomeScreenButtonProps & { className
export const FabricHomeScreen: React.FC<SplashScreenProps> = (props: SplashScreenProps) => {
const styles = useStyles();
const [openSampleDataImportDialog, setOpenSampleDataImportDialog] = React.useState(false);
const [selectedSampleDataConfiguration, setSelectedSampleDataConfiguration] =
React.useState<SampleDataConfiguration>(undefined);
const getSplashScreenButtons = (): JSX.Element => {
const buttons: FabricHomeScreenButtonProps[] = [
@@ -145,10 +168,30 @@ export const FabricHomeScreen: React.FC<SplashScreenProps> = (props: SplashScree
},
},
{
title: "Sample data",
description: "Automatically load sample data in your database",
title: "Sample Data",
description: "Load sample data in your database",
icon: <img src={CosmosDbBlackIcon} alt={"Azure Cosmos DB icon"} aria-hidden="true" />,
onClick: () => setOpenSampleDataImportDialog(true),
onClick: () => {
setSelectedSampleDataConfiguration({
databaseName: userContext.fabricContext?.databaseName,
newContainerName: "SampleData",
sampleDataFile: SampleDataFile.FABRIC_SAMPLE_DATA,
});
setOpenSampleDataImportDialog(true);
},
},
{
title: "Sample Vector Data",
description: "Load sample vector data in your database",
icon: <img src={AzureOpenAiIcon} alt={"Azure Open AI icon"} aria-hidden="true" />,
onClick: () => {
setSelectedSampleDataConfiguration({
databaseName: userContext.fabricContext?.databaseName,
newContainerName: "SampleVectorData",
sampleDataFile: SampleDataFile.FABRIC_SAMPLE_VECTOR_DATA,
});
setOpenSampleDataImportDialog(true);
},
},
{
title: "App development",
@@ -156,17 +199,25 @@ export const FabricHomeScreen: React.FC<SplashScreenProps> = (props: SplashScree
icon: <LinkMultipleRegular />,
onClick: () => window.open("https://aka.ms/cosmosdbfabricsdk", "_blank"),
},
{
title: "Sample Gallery",
description: "Get real-world end-to-end samples",
icon: <img src={GithubIcon} alt={"GitHub icon"} aria-hidden="true" />,
onClick: () => window.open("https://azurecosmosdb.github.io/gallery/?tags=example&tags=analytics", "_blank"),
},
];
return isFabricNativeReadOnly() ? (
<div className={styles.buttonsContainer}>
<FabricHomeScreenButton className={styles.single} {...buttons[2]} />
<FabricHomeScreenButton className={styles.single} {...buttons[3]} />
</div>
) : (
<div className={styles.buttonsContainer}>
<FabricHomeScreenButton className={styles.one} {...buttons[0]} />
<FabricHomeScreenButton className={styles.two} {...buttons[1]} />
<FabricHomeScreenButton className={styles.three} {...buttons[2]} />
<FabricHomeScreenButton className={styles.four} {...buttons[3]} />
<FabricHomeScreenButton className={styles.five} {...buttons[4]} />
</div>
);
};
@@ -179,7 +230,7 @@ export const FabricHomeScreen: React.FC<SplashScreenProps> = (props: SplashScree
open={openSampleDataImportDialog}
setOpen={setOpenSampleDataImportDialog}
explorer={props.explorer}
databaseName={userContext.fabricContext?.databaseName}
sampleDataConfiguration={selectedSampleDataConfiguration}
/>
<div className={styles.title} role="heading" aria-label={title} aria-level={1}>
{title}

View File

@@ -11,12 +11,10 @@ import {
tokens,
} from "@fluentui/react-components";
import Explorer from "Explorer/Explorer";
import { checkContainerExists, createContainer, importData } from "Explorer/SplashScreen/SampleUtil";
import { checkContainerExists, createContainer, importData, SampleDataFile } from "Explorer/SplashScreen/SampleUtil";
import React, { useEffect, useState } from "react";
import * as ViewModels from "../../Contracts/ViewModels";
const SAMPLE_DATA_CONTAINER_NAME = "SampleData";
const useStyles = makeStyles({
dialogContent: {
alignItems: "center",
@@ -24,6 +22,12 @@ const useStyles = makeStyles({
},
});
export interface SampleDataConfiguration {
databaseName: string;
newContainerName: string;
sampleDataFile: SampleDataFile;
}
/**
* This dialog:
* - creates a container
@@ -35,11 +39,11 @@ export const SampleDataImportDialog: React.FC<{
open: boolean;
setOpen: (open: boolean) => void;
explorer: Explorer;
databaseName: string;
sampleDataConfiguration: SampleDataConfiguration | undefined;
}> = (props) => {
const [status, setStatus] = useState<"idle" | "creating" | "importing" | "completed" | "error">("idle");
const [errorMessage, setErrorMessage] = useState<string | null>(null);
const containerName = SAMPLE_DATA_CONTAINER_NAME;
const containerName = props.sampleDataConfiguration?.newContainerName;
const [collection, setCollection] = useState<ViewModels.Collection>(undefined);
const styles = useStyles();
@@ -53,7 +57,7 @@ export const SampleDataImportDialog: React.FC<{
const handleStartImport = async (): Promise<void> => {
setStatus("creating");
const databaseName = props.databaseName;
const databaseName = props.sampleDataConfiguration.databaseName;
if (checkContainerExists(databaseName, containerName)) {
const msg = `The container "${containerName}" in database "${databaseName}" already exists. Please delete it and retry.`;
setStatus("error");
@@ -63,7 +67,12 @@ export const SampleDataImportDialog: React.FC<{
let collection;
try {
collection = await createContainer(databaseName, containerName, props.explorer);
collection = await createContainer(
databaseName,
containerName,
props.explorer,
props.sampleDataConfiguration.sampleDataFile,
);
} catch (error) {
setStatus("error");
setErrorMessage(`Failed to create container: ${error instanceof Error ? error.message : String(error)}`);
@@ -72,7 +81,7 @@ export const SampleDataImportDialog: React.FC<{
try {
setStatus("importing");
await importData(collection);
await importData(props.sampleDataConfiguration.sampleDataFile, collection);
setCollection(collection);
setStatus("completed");
} catch (error) {

View File

@@ -1,9 +1,13 @@
import { JSONObject } from "@azure/cosmos";
import { BackendDefaults } from "Common/Constants";
import { createCollection } from "Common/dataAccess/createCollection";
import Explorer from "Explorer/Explorer";
import { useDatabases } from "Explorer/useDatabases";
import { DEFAULT_FABRIC_NATIVE_CONTAINER_THROUGHPUT, isFabricNative } from "Platform/Fabric/FabricUtil";
import { Action, ActionModifiers } from "Shared/Telemetry/TelemetryConstants";
import * as DataModels from "../../Contracts/DataModels";
import * as ViewModels from "../../Contracts/ViewModels";
import * as TelemetryProcessor from "../../Shared/Telemetry/TelemetryProcessor";
/**
* Public for unit tests
@@ -26,12 +30,20 @@ const hasContainer = (
export const checkContainerExists = (databaseName: string, containerName: string) =>
hasContainer(databaseName, containerName, useDatabases.getState().databases);
export enum SampleDataFile {
COPILOT = "Copilot",
FABRIC_SAMPLE_DATA = "FabricSampleData",
FABRIC_SAMPLE_VECTOR_DATA = "FabricSampleVectorData",
}
export const createContainer = async (
databaseName: string,
containerName: string,
explorer: Explorer,
sampleDataFile: SampleDataFile,
): Promise<ViewModels.Collection> => {
const createRequest: DataModels.CreateCollectionParams = {
autoPilotMaxThroughput: isFabricNative() ? DEFAULT_FABRIC_NATIVE_CONTAINER_THROUGHPUT : undefined,
createNewDatabase: false,
collectionId: containerName,
databaseId: databaseName,
@@ -41,6 +53,44 @@ export const createContainer = async (
kind: "Hash",
version: BackendDefaults.partitionKeyVersion,
},
vectorEmbeddingPolicy:
sampleDataFile === SampleDataFile.FABRIC_SAMPLE_VECTOR_DATA
? {
vectorEmbeddings: [
{
path: "/descriptionVector",
dataType: "float32",
distanceFunction: "cosine",
dimensions: 512,
},
],
}
: undefined,
indexingPolicy:
sampleDataFile === SampleDataFile.FABRIC_SAMPLE_VECTOR_DATA
? {
automatic: true,
indexingMode: "consistent",
includedPaths: [
{
path: "/*",
},
],
excludedPaths: [
{
path: '/"_etag"/?',
},
],
fullTextIndexes: [],
vectorIndexes: [
{
path: "/descriptionVector",
type: "quantizedFlat",
quantizationByteSize: 64,
},
],
}
: undefined,
};
await createCollection(createRequest);
await explorer.refreshAllDatabases();
@@ -55,10 +105,39 @@ export const createContainer = async (
const SAMPLE_DATA_PARTITION_KEY = "category"; // This pkey is specifically set for queryCopilotSampleData.json below
export const importData = async (collection: ViewModels.Collection): Promise<void> => {
// TODO: keep same chunk as ContainerSampleGenerator
const dataFileContent = await import(
/* webpackChunkName: "queryCopilotSampleData" */ "../../../sampleData/queryCopilotSampleData.json"
);
await collection.bulkInsertDocuments(dataFileContent.data);
export const importData = async (sampleDataFile: SampleDataFile, collection: ViewModels.Collection): Promise<void> => {
let documents: JSONObject[] = undefined;
switch (sampleDataFile) {
case SampleDataFile.COPILOT:
documents = (
await import(/* webpackChunkName: "queryCopilotSampleData" */ "../../../sampleData/queryCopilotSampleData.json")
).data;
break;
case SampleDataFile.FABRIC_SAMPLE_DATA:
documents = (await import(/* webpackChunkName: "fabricSampleData" */ "../../../sampleData/fabricSampleData.json"))
.default;
break;
case SampleDataFile.FABRIC_SAMPLE_VECTOR_DATA:
documents = (
await import(
/* webpackChunkName: "fabricSampleDataVectors" */ "../../../sampleData/fabricSampleDataVectors.json"
)
).default;
break;
default:
throw new Error(`Unknown sample data file: ${sampleDataFile}`);
}
if (!documents) {
throw new Error(`Failed to load sample data file: ${sampleDataFile}`);
}
// Time it
const start = performance.now();
await collection.bulkInsertDocuments(documents);
const end = performance.now();
TelemetryProcessor.trace(Action.ImportSampleData, ActionModifiers.Success, {
documentsCount: documents.length,
durationMs: end - start,
sampleDataFile,
});
};