Cloud Integrations Tutorial - SeleniumBase for Rust

SeleniumBase for Rust can upload test artifacts (screenshots, logs, PDFs) to Amazon S3, Azure Blob Storage, and Google Cloud Storage. Each backend is behind an optional Cargo feature so the default build stays lightweight.

These types are uploaders, not test hooks: call upload_file after a test, for example on the screenshot a failure plugin saved. See Test Plugins for the hooks.

Info

What you will learn

Which Cargo features enable each cloud backend.
How to authenticate with AWS, Azure, and GCS.
How to upload files from a test.
How to run the included cloud_upload example.

Enabling backends#

Add the features you need, in Cargo.toml or on the command line:

[dependencies]seleniumbase-rs = { git = "https://github.com/MustCodeAl/seleniumbase-rs", features = ["s3", "azure", "gcp"] }

cargo run --features s3

Providers#

Feature: s3

Set the standard AWS environment variables:

export AWS_ACCESS_KEY_ID=AKIA...export AWS_SECRET_ACCESS_KEY=...export AWS_REGION=us-east-1   # optional fallback

MinIO and other S3-compatible stores are supported via AWS_ENDPOINT_URL:

export AWS_ENDPOINT_URL=https://minio.example.com

Upload from a test:

use seleniumbase_rs::plugins::s3_logging_plugin::S3LoggingPlugin;use std::path::Path;S3LoggingPlugin::upload_file(    "my-bucket",    "seleniumbase/logs/session.log",    Path::new("latest_logs/last_run.log"),    "us-east-1",).await?;

Feature: azure

Azure support uses a blob URL that already contains a SAS token with write permission:

export AZURE_BLOB_URL="https://myaccount.blob.core.windows.net/mycontainer/myblob?sv=...&sig=...&sp=rw"

Upload from a test:

use seleniumbase_rs::plugins::azure_logging_plugin::AzureLoggingPlugin;use std::path::Path;AzureLoggingPlugin::upload_file(Path::new("latest_logs/last_run.log")).await?;

Generate a SAS token with the Azure CLI:

az storage blob generate-sas \  --account-name myaccount \  --container-name mycontainer \  --name myblob \  --permissions rw \  --expiry 2027-12-31T00:00:00Z \  --https-only

Feature: gcp

GCS support uses an OAuth access token. The easiest way to obtain one is:

export GCS_ACCESS_TOKEN=$(gcloud auth print-access-token)

Upload from a test:

use seleniumbase_rs::plugins::gcp_logging_plugin::GcpLoggingPlugin;use std::path::Path;GcpLoggingPlugin::upload_file(    "my-bucket",    "seleniumbase/logs/session.log",    Path::new("latest_logs/last_run.log"),).await?;

Running the example#

cargo run --example cloud_upload --features s3,azure,gcp

The example writes a small log file to latest_logs/ and only performs an upload when the corresponding feature is enabled and its environment variables are set. A backend that is compiled in but not configured is reported as skipped, with the variables to set; the example fails only when an upload it attempted fails. Besides the credentials described above, it reads the bucket name from the environment, so nothing is hard-coded:

Feature	Variables the example reads
`s3`	`SB_S3_BUCKET`, `AWS_ACCESS_KEY_ID`, `AWS_SECRET_ACCESS_KEY`; optional `AWS_REGION` and `AWS_ENDPOINT_URL`
`azure`	`AZURE_BLOB_URL`
`gcp`	`SB_GCS_BUCKET`, `GCS_ACCESS_TOKEN`

Troubleshooting#

Upload function not found

Likely cause: Feature not enabled.

Fix: Add the matching feature in Cargo.toml.

S3 credentials error

Likely cause: Missing AWS_* variables.

Fix: Export AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY.

Azure SAS rejected

Likely cause: Token lacks write permission.

Fix: Regenerate with --permissions rw.

GCS token expired

Likely cause: GCS_ACCESS_TOKEN timed out.

Fix: Re-run gcloud auth print-access-token.