feature: databricks sync

This commit is contained in:
Scott Wilson
2025-02-10 18:38:27 -08:00
parent 4e48ab1eeb
commit 4de8888843
105 changed files with 1548 additions and 173 deletions
@@ -1,6 +1,7 @@
export enum AppConnection {
GitHub = "github",
AWS = "aws",
Databricks = "databricks",
GCP = "gcp",
AzureKeyVault = "azure-key-vault",
AzureAppConfiguration = "azure-app-configuration"
@@ -5,12 +5,17 @@ import { TAppConnectionServiceFactoryDep } from "@app/services/app-connection/ap
import { TAppConnection, TAppConnectionConfig } from "@app/services/app-connection/app-connection-types";
import {
AwsConnectionMethod,
getAwsAppConnectionListItem,
getAwsConnectionListItem,
validateAwsConnectionCredentials
} from "@app/services/app-connection/aws";
import {
DatabricksConnectionMethod,
getDatabricksConnectionListItem,
validateDatabricksConnectionCredentials
} from "@app/services/app-connection/databricks";
import {
GcpConnectionMethod,
getGcpAppConnectionListItem,
getGcpConnectionListItem,
validateGcpConnectionCredentials
} from "@app/services/app-connection/gcp";
import {
@@ -33,11 +38,12 @@ import {
export const listAppConnectionOptions = () => {
return [
getAwsAppConnectionListItem(),
getAwsConnectionListItem(),
getGitHubConnectionListItem(),
getGcpAppConnectionListItem(),
getGcpConnectionListItem(),
getAzureKeyVaultConnectionListItem(),
getAzureAppConfigurationConnectionListItem()
getAzureAppConfigurationConnectionListItem(),
getDatabricksConnectionListItem()
].sort((a, b) => a.name.localeCompare(b.name));
};
@@ -90,6 +96,8 @@ export const validateAppConnectionCredentials = async (
switch (app) {
case AppConnection.AWS:
return validateAwsConnectionCredentials(appConnection);
case AppConnection.Databricks:
return validateDatabricksConnectionCredentials(appConnection);
case AppConnection.GitHub:
return validateGitHubConnectionCredentials(appConnection);
case AppConnection.GCP:
@@ -118,6 +126,8 @@ export const getAppConnectionMethodName = (method: TAppConnection["method"]) =>
return "Assume Role";
case GcpConnectionMethod.ServiceAccountImpersonation:
return "Service Account Impersonation";
case DatabricksConnectionMethod.ServicePrincipal:
return "Service Principal";
default:
// eslint-disable-next-line @typescript-eslint/restrict-template-expressions
throw new Error(`Unhandled App Connection Method: ${method}`);
@@ -5,5 +5,6 @@ export const APP_CONNECTION_NAME_MAP: Record<AppConnection, string> = {
[AppConnection.GitHub]: "GitHub",
[AppConnection.GCP]: "GCP",
[AppConnection.AzureKeyVault]: "Azure Key Vault",
[AppConnection.AzureAppConfiguration]: "Azure App Configuration"
[AppConnection.AzureAppConfiguration]: "Azure App Configuration",
[AppConnection.Databricks]: "Databricks"
};
@@ -23,6 +23,8 @@ import {
TValidateAppConnectionCredentials
} from "@app/services/app-connection/app-connection-types";
import { ValidateAwsConnectionCredentialsSchema } from "@app/services/app-connection/aws";
import { ValidateDatabricksConnectionCredentialsSchema } from "@app/services/app-connection/databricks";
import { databricksConnectionService } from "@app/services/app-connection/databricks/databricks-connection-service";
import { ValidateGitHubConnectionCredentialsSchema } from "@app/services/app-connection/github";
import { githubConnectionService } from "@app/services/app-connection/github/github-connection-service";
import { TKmsServiceFactory } from "@app/services/kms/kms-service";
@@ -46,7 +48,8 @@ const VALIDATE_APP_CONNECTION_CREDENTIALS_MAP: Record<AppConnection, TValidateAp
[AppConnection.GitHub]: ValidateGitHubConnectionCredentialsSchema,
[AppConnection.GCP]: ValidateGcpConnectionCredentialsSchema,
[AppConnection.AzureKeyVault]: ValidateAzureKeyVaultConnectionCredentialsSchema,
[AppConnection.AzureAppConfiguration]: ValidateAzureAppConfigurationConnectionCredentialsSchema
[AppConnection.AzureAppConfiguration]: ValidateAzureAppConfigurationConnectionCredentialsSchema,
[AppConnection.Databricks]: ValidateDatabricksConnectionCredentialsSchema
};
export const appConnectionServiceFactory = ({
@@ -365,6 +368,7 @@ export const appConnectionServiceFactory = ({
connectAppConnectionById,
listAvailableAppConnectionsForUser,
github: githubConnectionService(connectAppConnectionById),
gcp: gcpConnectionService(connectAppConnectionById)
gcp: gcpConnectionService(connectAppConnectionById),
databricks: databricksConnectionService(connectAppConnectionById, appConnectionDAL, kmsService)
};
};
@@ -4,6 +4,12 @@ import {
TAwsConnectionInput,
TValidateAwsConnectionCredentials
} from "@app/services/app-connection/aws";
import {
TDatabricksConnection,
TDatabricksConnectionConfig,
TDatabricksConnectionInput,
TValidateDatabricksConnectionCredentials
} from "@app/services/app-connection/databricks";
import {
TGitHubConnection,
TGitHubConnectionConfig,
@@ -31,6 +37,7 @@ export type TAppConnection = { id: string } & (
| TGcpConnection
| TAzureKeyVaultConnection
| TAzureAppConfigurationConnection
| TDatabricksConnection
);
export type TAppConnectionInput = { id: string } & (
@@ -39,6 +46,7 @@ export type TAppConnectionInput = { id: string } & (
| TGcpConnectionInput
| TAzureKeyVaultConnectionInput
| TAzureAppConfigurationConnectionInput
| TDatabricksConnectionInput
);
export type TCreateAppConnectionDTO = Pick<
@@ -55,11 +63,13 @@ export type TAppConnectionConfig =
| TGitHubConnectionConfig
| TGcpConnectionConfig
| TAzureKeyVaultConnectionConfig
| TAzureAppConfigurationConnectionConfig;
| TAzureAppConfigurationConnectionConfig
| TDatabricksConnectionConfig;
export type TValidateAppConnectionCredentials =
| TValidateAwsConnectionCredentials
| TValidateGitHubConnectionCredentials
| TValidateGcpConnectionCredentials
| TValidateAzureKeyVaultConnectionCredentials
| TValidateAzureAppConfigurationConnectionCredentials;
| TValidateAzureAppConfigurationConnectionCredentials
| TValidateDatabricksConnectionCredentials;
@@ -9,7 +9,7 @@ import { AppConnection, AWSRegion } from "@app/services/app-connection/app-conne
import { AwsConnectionMethod } from "./aws-connection-enums";
import { TAwsConnectionConfig } from "./aws-connection-types";
export const getAwsAppConnectionListItem = () => {
export const getAwsConnectionListItem = () => {
const { INF_APP_CONNECTION_AWS_ACCESS_KEY_ID } = getConfig();
return {
@@ -0,0 +1,3 @@
export enum DatabricksConnectionMethod {
ServicePrincipal = "service-principal"
}
@@ -0,0 +1,92 @@
import { request } from "@app/lib/config/request";
import { BadRequestError } from "@app/lib/errors";
import { removeTrailingSlash } from "@app/lib/fn";
import { TAppConnectionDALFactory } from "@app/services/app-connection/app-connection-dal";
import { AppConnection } from "@app/services/app-connection/app-connection-enums";
import { encryptAppConnectionCredentials } from "@app/services/app-connection/app-connection-fns";
import { TKmsServiceFactory } from "@app/services/kms/kms-service";
import { DatabricksConnectionMethod } from "./databricks-connection-enums";
import {
TAuthorizeDatabricksConnection,
TDatabricksConnection,
TDatabricksConnectionConfig
} from "./databricks-connection-types";
export const getDatabricksConnectionListItem = () => {
return {
name: "Databricks" as const,
app: AppConnection.Databricks as const,
methods: Object.values(DatabricksConnectionMethod) as [DatabricksConnectionMethod.ServicePrincipal]
};
};
const authorizeDatabricksConnection = async ({
clientId,
clientSecret,
workspaceUrl
}: Pick<TDatabricksConnection["credentials"], "workspaceUrl" | "clientId" | "clientSecret">) => {
const { data } = await request.post<TAuthorizeDatabricksConnection>(
`${removeTrailingSlash(workspaceUrl)}/oidc/v1/token`,
"grant_type=client_credentials&scope=all-apis",
{
auth: {
username: clientId,
password: clientSecret
},
headers: {
"Content-Type": "application/x-www-form-urlencoded"
}
}
);
return { accessToken: data.access_token, expiresAt: data.expires_in * 1000 + Date.now() };
};
export const getDatabricksConnectionAccessToken = async (
{ id, orgId, credentials }: TDatabricksConnection,
appConnectionDAL: Pick<TAppConnectionDALFactory, "updateById">,
kmsService: Pick<TKmsServiceFactory, "createCipherPairWithDataKey">
) => {
const { clientSecret, clientId, workspaceUrl, accessToken, expiresAt } = credentials;
// get new token if less than 10 minutes from expiry
if (Date.now() < expiresAt - 10_000) {
return accessToken;
}
const authData = await authorizeDatabricksConnection({ clientId, clientSecret, workspaceUrl });
const updatedCredentials: TDatabricksConnection["credentials"] = {
...credentials,
...authData
};
const encryptedCredentials = await encryptAppConnectionCredentials({
credentials: updatedCredentials,
orgId,
kmsService
});
await appConnectionDAL.updateById(id, { encryptedCredentials });
return authData.accessToken;
};
export const validateDatabricksConnectionCredentials = async (appConnection: TDatabricksConnectionConfig) => {
const { credentials } = appConnection;
try {
const { accessToken, expiresAt } = await authorizeDatabricksConnection(appConnection.credentials);
return {
...credentials,
accessToken,
expiresAt
};
} catch (e: unknown) {
throw new BadRequestError({
message: `Unable to validate connection - verify credentials`
});
}
};
@@ -0,0 +1,77 @@
import { z } from "zod";
import { AppConnections } from "@app/lib/api-docs";
import { AppConnection } from "@app/services/app-connection/app-connection-enums";
import {
BaseAppConnectionSchema,
GenericCreateAppConnectionFieldsSchema,
GenericUpdateAppConnectionFieldsSchema
} from "@app/services/app-connection/app-connection-schemas";
import { DatabricksConnectionMethod } from "./databricks-connection-enums";
export const DatabricksConnectionServicePrincipalInputCredentialsSchema = z.object({
clientId: z.string().trim().min(1, "Client ID required"),
clientSecret: z.string().trim().min(1, "Client Secret required"),
workspaceUrl: z.string().trim().url().min(1, "Workspace URL required")
});
export const DatabricksConnectionServicePrincipalOutputCredentialsSchema = z
.object({
accessToken: z.string(),
expiresAt: z.number()
})
.merge(DatabricksConnectionServicePrincipalInputCredentialsSchema);
const BaseDatabricksConnectionSchema = BaseAppConnectionSchema.extend({ app: z.literal(AppConnection.Databricks) });
export const DatabricksConnectionSchema = z.intersection(
BaseDatabricksConnectionSchema,
z.discriminatedUnion("method", [
z.object({
method: z.literal(DatabricksConnectionMethod.ServicePrincipal),
credentials: DatabricksConnectionServicePrincipalOutputCredentialsSchema
})
])
);
export const SanitizedDatabricksConnectionSchema = z.discriminatedUnion("method", [
BaseDatabricksConnectionSchema.extend({
method: z.literal(DatabricksConnectionMethod.ServicePrincipal),
credentials: DatabricksConnectionServicePrincipalOutputCredentialsSchema.pick({
clientId: true,
workspaceUrl: true
})
})
]);
export const ValidateDatabricksConnectionCredentialsSchema = z.discriminatedUnion("method", [
z.object({
method: z
.literal(DatabricksConnectionMethod.ServicePrincipal)
.describe(AppConnections?.CREATE(AppConnection.Databricks).method),
credentials: DatabricksConnectionServicePrincipalInputCredentialsSchema.describe(
AppConnections.CREATE(AppConnection.Databricks).credentials
)
})
]);
export const CreateDatabricksConnectionSchema = ValidateDatabricksConnectionCredentialsSchema.and(
GenericCreateAppConnectionFieldsSchema(AppConnection.Databricks)
);
export const UpdateDatabricksConnectionSchema = z
.object({
credentials: DatabricksConnectionServicePrincipalInputCredentialsSchema.optional().describe(
AppConnections.UPDATE(AppConnection.Databricks).credentials
)
})
.and(GenericUpdateAppConnectionFieldsSchema(AppConnection.Databricks));
export const DatabricksConnectionListItemSchema = z.object({
name: z.literal("Databricks"),
app: z.literal(AppConnection.Databricks),
// the below is preferable but currently breaks with our zod to json schema parser
// methods: z.tuple([z.literal(AwsConnectionMethod.ServicePrincipal), z.literal(AwsConnectionMethod.AccessKey)]),
methods: z.nativeEnum(DatabricksConnectionMethod).array()
});
@@ -0,0 +1,60 @@
import { request } from "@app/lib/config/request";
import { removeTrailingSlash } from "@app/lib/fn";
import { OrgServiceActor } from "@app/lib/types";
import { TAppConnectionDALFactory } from "@app/services/app-connection/app-connection-dal";
import { AppConnection } from "@app/services/app-connection/app-connection-enums";
import { getDatabricksConnectionAccessToken } from "@app/services/app-connection/databricks/databricks-connection-fns";
import {
TDatabricksConnection,
TDatabricksListSecretScopesResponse
} from "@app/services/app-connection/databricks/databricks-connection-types";
import { TKmsServiceFactory } from "@app/services/kms/kms-service";
type TGetAppConnectionFunc = (
app: AppConnection,
connectionId: string,
actor: OrgServiceActor
) => Promise<TDatabricksConnection>;
const listDatabricksSecretScopes = async (
appConnection: TDatabricksConnection,
appConnectionDAL: Pick<TAppConnectionDALFactory, "updateById">,
kmsService: Pick<TKmsServiceFactory, "createCipherPairWithDataKey">
) => {
const {
credentials: { workspaceUrl }
} = appConnection;
const accessToken = await getDatabricksConnectionAccessToken(appConnection, appConnectionDAL, kmsService);
const { data } = await request.get<TDatabricksListSecretScopesResponse>(
`${removeTrailingSlash(workspaceUrl)}/api/2.0/secrets/scopes/list`,
{
headers: {
Authorization: `Bearer ${accessToken}`,
"Accept-Encoding": "application/json"
}
}
);
// not present in response if no scopes exists
return data.scopes ?? [];
};
export const databricksConnectionService = (
getAppConnection: TGetAppConnectionFunc,
appConnectionDAL: Pick<TAppConnectionDALFactory, "updateById">,
kmsService: Pick<TKmsServiceFactory, "createCipherPairWithDataKey">
) => {
const listSecretScopes = async (connectionId: string, actor: OrgServiceActor) => {
const appConnection = await getAppConnection(AppConnection.Databricks, connectionId, actor);
const secretScopes = await listDatabricksSecretScopes(appConnection, appConnectionDAL, kmsService);
return secretScopes;
};
return {
listSecretScopes
};
};
@@ -0,0 +1,36 @@
import { z } from "zod";
import { DiscriminativePick } from "@app/lib/types";
import { AppConnection } from "@app/services/app-connection/app-connection-enums";
import {
CreateDatabricksConnectionSchema,
DatabricksConnectionSchema,
ValidateDatabricksConnectionCredentialsSchema
} from "./databricks-connection-schemas";
export type TDatabricksConnection = z.infer<typeof DatabricksConnectionSchema>;
export type TDatabricksConnectionInput = z.infer<typeof CreateDatabricksConnectionSchema> & {
app: AppConnection.Databricks;
};
export type TValidateDatabricksConnectionCredentials = typeof ValidateDatabricksConnectionCredentialsSchema;
export type TDatabricksConnectionConfig = DiscriminativePick<
TDatabricksConnection,
"method" | "app" | "credentials"
> & {
orgId: string;
};
export type TAuthorizeDatabricksConnection = {
access_token: string;
scope: string;
token_type: string;
expires_in: number;
};
export type TDatabricksListSecretScopesResponse = {
scopes?: { name: string; backend_type: string; keyvault_metadata: { resource_id: string; dns_name: string } }[];
};
@@ -0,0 +1,4 @@
export * from "./databricks-connection-enums";
export * from "./databricks-connection-fns";
export * from "./databricks-connection-schemas";
export * from "./databricks-connection-types";
@@ -17,7 +17,7 @@ import {
TGcpConnectionConfig
} from "./gcp-connection-types";
export const getGcpAppConnectionListItem = () => {
export const getGcpConnectionListItem = () => {
return {
name: "GCP" as const,
app: AppConnection.GCP as const,