Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 6 additions & 2 deletions bottlecap/src/bin/bottlecap/main.rs
Original file line number Diff line number Diff line change
Expand Up @@ -162,7 +162,7 @@ fn build_function_arn(account_id: &str, region: &str, function_name: &str) -> St

#[tokio::main]
async fn main() -> Result<()> {
let (aws_config, config) = load_configs();
let (mut aws_config, config) = load_configs();

enable_logging_subsystem(&config);
let client = reqwest::Client::builder().no_proxy().build().map_err(|e| {
Expand All @@ -176,7 +176,7 @@ async fn main() -> Result<()> {
.await
.map_err(|e| Error::new(std::io::ErrorKind::InvalidData, e.to_string()))?;

if let Some(resolved_api_key) = resolve_secrets(Arc::clone(&config), &aws_config).await {
if let Some(resolved_api_key) = resolve_secrets(Arc::clone(&config), &mut aws_config).await {
match extension_loop_active(&aws_config, &config, &client, &r, resolved_api_key).await {
Ok(()) => {
debug!("Extension loop completed successfully");
Expand All @@ -202,6 +202,10 @@ fn load_configs() -> (AwsConfig, Arc<Config>) {
aws_access_key_id: env::var("AWS_ACCESS_KEY_ID").unwrap_or_default(),
aws_secret_access_key: env::var("AWS_SECRET_ACCESS_KEY").unwrap_or_default(),
aws_session_token: env::var("AWS_SESSION_TOKEN").unwrap_or_default(),
aws_container_credentials_full_uri: env::var("AWS_CONTAINER_CREDENTIALS_FULL_URI")
.unwrap_or_default(),
aws_container_authorization_token: env::var("AWS_CONTAINER_AUTHORIZATION_TOKEN")
.unwrap_or_default(),
function_name: env::var("AWS_LAMBDA_FUNCTION_NAME").unwrap_or_default(),
sandbox_init_time: Instant::now(),
};
Expand Down
2 changes: 2 additions & 0 deletions bottlecap/src/config/mod.rs
Original file line number Diff line number Diff line change
Expand Up @@ -334,6 +334,8 @@ pub struct AwsConfig {
pub aws_session_token: String,
pub function_name: String,
pub sandbox_init_time: Instant,
pub aws_container_credentials_full_uri: String,
pub aws_container_authorization_token: String,
}

#[must_use]
Expand Down
2 changes: 2 additions & 0 deletions bottlecap/src/lifecycle/invocation/processor.rs
Original file line number Diff line number Diff line change
Expand Up @@ -632,6 +632,8 @@ mod tests {
aws_access_key_id: "***".into(),
aws_secret_access_key: "***".into(),
aws_session_token: "***".into(),
aws_container_credentials_full_uri: "***".into(),
aws_container_authorization_token: "***".into(),
function_name: "test-function".into(),
sandbox_init_time: Instant::now(),
};
Expand Down
51 changes: 49 additions & 2 deletions bottlecap/src/secrets/decrypt.rs
Original file line number Diff line number Diff line change
Expand Up @@ -12,7 +12,7 @@ use std::time::Instant;
use tracing::debug;
use tracing::error;

pub async fn resolve_secrets(config: Arc<Config>, aws_config: &AwsConfig) -> Option<String> {
pub async fn resolve_secrets(config: Arc<Config>, aws_config: &mut AwsConfig) -> Option<String> {
let api_key_candidate =
if !config.api_key_secret_arn.is_empty() || !config.kms_api_key.is_empty() {
let before_decrypt = Instant::now();
Expand All @@ -25,6 +25,33 @@ pub async fn resolve_secrets(config: Arc<Config>, aws_config: &AwsConfig) -> Opt
}
};

if aws_config.aws_secret_access_key.is_empty()
&& aws_config.aws_access_key_id.is_empty()
&& !aws_config.aws_container_credentials_full_uri.is_empty()
&& !aws_config.aws_container_authorization_token.is_empty()
{
// We're in Snap Start
let credentials = match get_snapstart_credentials(aws_config, &client).await {
Ok(credentials) => credentials,
Err(err) => {
error!("Error getting Snap Start credentials: {}", err);
return None;
}
};
aws_config.aws_access_key_id = credentials["AccessKeyId"]
.as_str()
.unwrap_or_default()
.to_string();
aws_config.aws_secret_access_key = credentials["SecretAccessKey"]
.as_str()
.unwrap_or_default()
.to_string();
aws_config.aws_session_token = credentials["Token"]
.as_str()
.unwrap_or_default()
.to_string();
}

let decrypted_key = if config.kms_api_key.is_empty() {
decrypt_aws_sm(&client, config.api_key_secret_arn.clone(), aws_config).await
} else {
Expand Down Expand Up @@ -146,6 +173,24 @@ async fn decrypt_aws_sm(
}
}

async fn get_snapstart_credentials(
aws_config: &AwsConfig,
client: &Client,
) -> Result<Value, Box<dyn std::error::Error>> {
let mut headers = HeaderMap::new();
headers.insert(
"Authorization",
HeaderValue::from_str(&aws_config.aws_container_authorization_token)?,
);

let req = client
.get(&aws_config.aws_container_credentials_full_uri)
.headers(headers);
let body = req.send().await?.text().await?;
let v: Value = serde_json::from_str(&body)?;
Ok(v)
}

async fn request(
json_body: &Value,
headers: HeaderMap,
Expand Down Expand Up @@ -283,11 +328,13 @@ mod tests {
&NaiveDateTime::parse_from_str("2024-05-30 09:10:11", "%Y-%m-%d %H:%M:%S").unwrap(),
);
let headers = build_get_secret_signed_headers(
&AwsConfig {
&AwsConfig{
region: "us-east-1".to_string(),
aws_access_key_id: "AKIDEXAMPLE".to_string(),
aws_secret_access_key: "wJalrXUtnFEMI/K7MDENG+bPxRfiCYEXAMPLEKEY".to_string(),
aws_session_token: "AQoDYXdzEJr...<remainder of session token>".to_string(),
aws_container_authorization_token: String::new(),
aws_container_credentials_full_uri: String::new(),
function_name: "arn:some-function".to_string(),
sandbox_init_time: Instant::now(),
},
Expand Down
24 changes: 24 additions & 0 deletions bottlecap/src/telemetry/events.rs
Original file line number Diff line number Diff line change
Expand Up @@ -125,6 +125,30 @@ pub enum TelemetryRecord {
/// Total size of the dropped records
dropped_bytes: u64,
},

/// Snapstart
#[serde(rename = "platform.restoreStart", rename_all = "camelCase")]
PlatformRestoreStart {
// function name and function version are here
// but we don't care about those
// https://docs.aws.amazon.com/lambda/latest/dg/telemetry-schema-reference.html#platform-restoreStart
// runtime version may be nice
},

#[serde(rename = "platform.restoreReport", rename_all = "camelCase")]
PlatformRestoreReport {
/// Status of the invocation
status: Status,
/// When unsuccessful, the `error_type` describes what kind of error occurred
error_type: Option<String>,
},
#[serde(rename = "platform.restoreRuntimeDone", rename_all = "camelCase")]
PlatformRestoreRuntimeDone {
Comment on lines +131 to +146
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Maybe we could provide some snapstart visibility eventually? seems easy enough

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think we could report the metric, yeah

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Even create a span

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why? It happens up to 2 weeks before an invocation occurs

/// Status of the invocation
status: Status,
/// When unsuccessful, the `error_type` describes what kind of error occurred
error_type: Option<String>,
},
}

/// Type of Initialization
Expand Down
9 changes: 5 additions & 4 deletions bottlecap/src/telemetry/listener.rs
Original file line number Diff line number Diff line change
Expand Up @@ -6,7 +6,7 @@ use tokio::sync::mpsc::Sender;

use hyper::service::{make_service_fn, service_fn};
use hyper::{Body, Request, Response, Server};
use tracing::{debug, error};
use tracing::debug;

#[allow(clippy::module_name_repetitions)]
#[derive(Debug, Clone, Copy)]
Expand Down Expand Up @@ -58,9 +58,10 @@ impl TelemetryListener {
Err(e) => {
// If we can't parse the event, we will receive it again in a new batch
// causing an infinite loop and resource contention.
// Instead, log it as fatal and move on.
// This will result in a dropped payload.
error!("[FATAL] Failed to parse telemetry events: {:?}", e);
// Instead, log it and move on.
// This will result in a dropped payload, but may be from
// events we haven't added support for yet
debug!("Failed to parse telemetry events: {:?}", e);
return Ok(Response::builder()
.status(hyper::StatusCode::OK)
.body(Body::from("Failed to parse telemetry events"))
Expand Down
Loading