Subscribe

UiPath AI Center

UiPath AI Center

Object storage

An S3 API-compatible interface is required by the AI Center installation for object storage.
Make sure that the buckets described below are created before proceeding with the AI Center installation.

Bucket

Recommended size

Description

aifabric-staging

5 GB

Used to store customer-uploaded ML model packages.

ml-model-files

100 GB

Used to store AI Center-validated ML model packages copied from aifabric-staging.

train-data

250 GB

Used to store training-related data.

datamanager

100 GB

To store documents used in data labeling

The following Cross-Origin Resource Sharing (CORS) policy must be set on all three buckets:

<CORSConfiguration>
 <CORSRule>
   <AllowedOrigin>*</AllowedOrigin>
   <AllowedMethod>PUT</AllowedMethod>
   <AllowedMethod>POST</AllowedMethod>
   <AllowedMethod>DELETE</AllowedMethod>
   <AllowedHeader>*</AllowedHeader>
  <MaxAgeSeconds>3000</MaxAgeSeconds>
  <ExposeHeader>x-amz-server-side-encryption</ExposeHeader>
  <ExposeHeader>x-amz-request-id</ExposeHeader>
  <ExposeHeader>x-amz-id-2</ExposeHeader>
 </CORSRule>
</CORSConfiguration>

Optionally, you can provide internal access details during the S3 storage configuration to provide faster access for requests from inside kubernetes workloads, if available.

Input parameter

Required

Description

Default

global.storage.accessKey

Yes

Access key for S3-compatible storage.

global.storage.secretKey

Yes

Secret key for S3-compatible storage.

global.storage.externalHost

Yes

Hostname for S3-compatible storage.

global.storage.externalPort

No

Port for S3-compatible storage.

443

global.storageCert

No

Certificate for HTTPS connection to storage.

global.storage.bucket.mlPackage.uploaded

No

Storage bucket to upload non-validated ML packages.

aifabric-staging

global.storage.bucket.mlPackage.validated

No

Storage bucket to transfer validated ML packages from non-validated ML packages storage bucket.

ml-model-files

global.storage.bucket.trainerPackage

No

Storage bucket to store training-related data and datasets.

train-data

global.storage.bucket.datamanager

No

Storage bucket to store data-labeling related documents

datamanager

Updated 2 months ago

Object storage


Suggested Edits are limited on API Reference Pages

You can only suggest edits to Markdown body content, but not to the API spec.