Support AWS_ENDPOINT_URL

.devcontainer/.env

@@ -2,10 +2,11 @@
 AWS_ACCESS_KEY_ID=minioadmin
 AWS_SECRET_ACCESS_KEY=minioadmin
 AWS_REGION=us-east-1
+AWS_ENDPOINT_URL=http://localhost:9000
+AWS_ALLOW_HTTP=true
 AWS_S3_TEST_BUCKET=testbucket
 MINIO_ROOT_USER=minioadmin
 MINIO_ROOT_PASSWORD=minioadmin
 
 # Others
 RUST_TEST_THREADS=1
-PG_PARQUET_TEST=true

.devcontainer/devcontainer.json

@@ -3,7 +3,6 @@
     "dockerComposeFile": "docker-compose.yml",
     "service": "app",
     "workspaceFolder": "/workspace",
-    "postStartCommand": "bash .devcontainer/create-test-buckets.sh",
     "postAttachCommand": "sudo chown -R rust /workspace",
     "customizations": {
         "vscode": {

.devcontainer/docker-compose.yml

@@ -23,10 +23,12 @@ services:
     env_file:
       - .env
     network_mode: host
-    command: server /data
+    entrypoint: "./entrypoint.sh"
     restart: unless-stopped
     healthcheck:
       test: ["CMD", "curl", "http://localhost:9000"]
       interval: 6s
       timeout: 2s
       retries: 3
+    volumes:
+      - ./minio-entrypoint.sh:/entrypoint.sh

.devcontainer/minio-entrypoint.sh

@@ -0,0 +1,20 @@
+#!/bin/bash
+
+trap "echo 'Caught termination signal. Exiting...'; exit 0" SIGINT SIGTERM
+
+minio server /data &
+
+minio_pid=$!
+
+while ! curl $AWS_ENDPOINT_URL; do
+    echo "Waiting for $AWS_ENDPOINT_URL..."
+    sleep 1
+done
+
+# set access key and secret key
+mc alias set local $AWS_ENDPOINT_URL $MINIO_ROOT_USER $MINIO_ROOT_PASSWORD
+
+# create bucket
+mc mb local/$AWS_S3_TEST_BUCKET
+
+wait $minio_pid

.github/workflows/ci.yml

@@ -107,15 +107,19 @@ jobs:
 
       - name: Start Minio for s3 emulator tests
         run: |
-          docker run -d --env-file .devcontainer/.env -p 9000:9000 minio/minio server /data
-
-          while ! nc -z localhost 9000; do
-              echo "Waiting for localhost:9000..."
-              sleep 1
+          docker run -d \
+            --env-file .devcontainer/.env \
+            -p 9000:9000 \
+            --entrypoint "./entrypoint.sh" \
+            --volume ./.devcontainer/minio-entrypoint.sh:/entrypoint.sh \
+            --name miniocontainer \
+            minio/minio
+
+          while ! curl $AWS_ENDPOINT_URL; do
+            echo "Waiting for $AWS_ENDPOINT_URL..."
+            sleep 1
           done
 
-          aws --endpoint-url http://localhost:9000 s3 mb s3://$AWS_S3_TEST_BUCKET
-
       - name: Run tests
         run: |
           # Run tests with coverage tool

Cargo.toml

@@ -23,8 +23,9 @@ pg_test = []
 arrow = {version = "53", default-features = false}
 arrow-cast = {version = "53", default-features = false}
 arrow-schema = {version = "53", default-features = false}
-aws-config = { version = "1.5", default-features = false, features = ["rustls"]}
-aws-credential-types = {version = "1.2", default-features = false}
+aws-config = { version = "1", default-features = false, features = ["rustls"]}
+aws-credential-types = {version = "1", default-features = false}
+aws-sdk-sts = "1"
 futures = "0.3"
 object_store = {version = "0.11", default-features = false, features = ["aws"]}
 once_cell = "1"

README.md

@@ -174,10 +174,14 @@ region = eu-central-1
 Alternatively, you can use the following environment variables when starting postgres to configure the S3 client:
 - `AWS_ACCESS_KEY_ID`: the access key ID of the AWS account
 - `AWS_SECRET_ACCESS_KEY`: the secret access key of the AWS account
+- `AWS_SESSION_TOKEN`: the session token for the AWS account
 - `AWS_REGION`: the default region of the AWS account
-- `AWS_SHARED_CREDENTIALS_FILE`: an alternative location for the credentials file
-- `AWS_CONFIG_FILE`: an alternative location for the config file
-- `AWS_PROFILE`: the name of the profile from the credentials and config file (default profile name is `default`)
+- `AWS_ENDPOINT_URL`: the endpoint
+- `AWS_SHARED_CREDENTIALS_FILE`: an alternative location for the credentials file **(only via environment variables)**
+- `AWS_CONFIG_FILE`: an alternative location for the config file **(only via environment variables)**
+- `AWS_PROFILE`: the name of the profile from the credentials and config file (default profile name is `default`) **(only via environment variables)**
+- `AWS_ALLOW_HTTP`: allows http endpoints **(only via environment variables)**
+
 
 > [!NOTE]
 > To be able to write into a object store location, you need to grant `parquet_object_store_write` role to your current postgres user.

src/arrow_parquet/uri_utils.rs

@@ -1,11 +1,7 @@
 use std::{sync::Arc, sync::LazyLock};
 
 use arrow::datatypes::SchemaRef;
-use aws_config::{
-    environment::{EnvironmentVariableCredentialsProvider, EnvironmentVariableRegionProvider},
-    meta::{credentials::CredentialsProviderChain, region::RegionProviderChain},
-    profile::{ProfileFileCredentialsProvider, ProfileFileRegionProvider},
-};
+use aws_config::BehaviorVersion;
 use aws_credential_types::provider::ProvideCredentials;
 use object_store::{
     aws::{AmazonS3, AmazonS3Builder},
@@ -92,48 +88,40 @@
 async fn get_s3_object_store(bucket_name: &str) -> AmazonS3 {
     let mut aws_s3_builder = AmazonS3Builder::new().with_bucket_name(bucket_name);
 
-    let is_test_running = std::env::var("PG_PARQUET_TEST").is_ok();
-
-    if is_test_running {
-        // use minio for testing
-        aws_s3_builder = aws_s3_builder.with_endpoint("http://localhost:9000");
-        aws_s3_builder = aws_s3_builder.with_allow_http(true);
-    }
+    // AWS_ALLOW_HTTP
+    if let Ok(aws_allow_http) = std::env::var("AWS_ALLOW_HTTP") {
+        aws_s3_builder = aws_s3_builder
+            .with_allow_http(aws_allow_http.parse().unwrap_or_else(|e| panic!("{}", e)));
+    };
 
-    let aws_profile_name = std::env::var("AWS_PROFILE").unwrap_or("default".to_string());
+    // first tries to load the profile files from the environment variables and then from the profile
+    let sdk_config = aws_config::defaults(BehaviorVersion::v2024_03_28())
+        .load()
+        .await;
 
-    let region_provider = RegionProviderChain::first_try(EnvironmentVariableRegionProvider::new())
-        .or_else(
-            ProfileFileRegionProvider::builder()
-                .profile_name(aws_profile_name.clone())
-                .build(),
-        );
+    if let Some(credential_provider) = sdk_config.credentials_provider() {
+        if let Ok(credentials) = credential_provider.provide_credentials().await {
+            // AWS_ACCESS_KEY_ID
+            aws_s3_builder = aws_s3_builder.with_access_key_id(credentials.access_key_id());
 
-    let region = region_provider.region().await;
+            // AWS_SECRET_ACCESS_KEY
+            aws_s3_builder = aws_s3_builder.with_secret_access_key(credentials.secret_access_key());
 
-    if let Some(region) = region {
-        aws_s3_builder = aws_s3_builder.with_region(region.to_string());
+            if let Some(token) = credentials.session_token() {
+                // AWS_SESSION_TOKEN
+                aws_s3_builder = aws_s3_builder.with_token(token);
+            }
+        }
     }
 
-    let credential_provider = CredentialsProviderChain::first_try(
-        "Environment",
-        EnvironmentVariableCredentialsProvider::new(),
-    )
-    .or_else(
-        "Profile",
-        ProfileFileCredentialsProvider::builder()
-            .profile_name(aws_profile_name)
-            .build(),
-    );
-
-    if let Ok(credentials) = credential_provider.provide_credentials().await {
-        aws_s3_builder = aws_s3_builder.with_access_key_id(credentials.access_key_id());
-
-        aws_s3_builder = aws_s3_builder.with_secret_access_key(credentials.secret_access_key());
+    // AWS_ENDPOINT_URL
+    if let Some(aws_endpoint_url) = sdk_config.endpoint_url() {
+        aws_s3_builder = aws_s3_builder.with_endpoint(aws_endpoint_url);
+    }
 
-        if let Some(token) = credentials.session_token() {
-            aws_s3_builder = aws_s3_builder.with_token(token);
-        }
+    // AWS_REGION
+    if let Some(aws_region) = sdk_config.region() {
+        aws_s3_builder = aws_s3_builder.with_region(aws_region.as_ref());
     }
 
     aws_s3_builder.build().unwrap_or_else(|e| panic!("{}", e))

src/lib.rs

@@ -5,6 +5,7 @@ use pgrx::{prelude::*, GucContext, GucFlags, GucRegistry};
 mod arrow_parquet;
 mod parquet_copy_hook;
 mod parquet_udfs;
+#[cfg(any(test, feature = "pg_test"))]
 mod pgrx_tests;
 mod pgrx_utils;
 mod type_compat;