1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522
// Copyright (c) 2021-2024 Espresso Systems (espressosys.com)
// This file is part of the HotShot repository.
// You should have received a copy of the MIT License
// along with the HotShot repository. If not, see <https://mit-license.org/>.
use std::{net::SocketAddr, time::Duration};
use clap::Parser;
use futures::{Future, FutureExt};
use hotshot_types::{
network::{NetworkConfig, NetworkConfigSource},
traits::signature_key::SignatureKey,
PeerConfig, ValidatorConfig,
};
use libp2p_identity::PeerId;
use multiaddr::Multiaddr;
use surf_disco::{error::ClientError, Client};
use tide_disco::Url;
use tokio::time::sleep;
use tracing::{info, instrument};
use vbs::BinarySerializer;
use crate::OrchestratorVersion;
/// Holds the client connection to the orchestrator
pub struct OrchestratorClient {
/// the client
pub client: surf_disco::Client<ClientError, OrchestratorVersion>,
}
/// Struct describing a benchmark result
#[derive(serde::Serialize, serde::Deserialize, Clone, Debug, Default, PartialEq)]
pub struct BenchResults {
/// Whether it's partial collected results
pub partial_results: String,
/// The average latency of the transactions
pub avg_latency_in_sec: i64,
/// The number of transactions that were latency measured
pub num_latency: i64,
/// The minimum latency of the transactions
pub minimum_latency_in_sec: i64,
/// The maximum latency of the transactions
pub maximum_latency_in_sec: i64,
/// The throughput of the consensus protocol = number of transactions committed per second * transaction size in bytes
pub throughput_bytes_per_sec: u64,
/// The number of transactions committed during benchmarking
pub total_transactions_committed: u64,
/// The size of each transaction in bytes
pub transaction_size_in_bytes: u64,
/// The total time elapsed for benchmarking
pub total_time_elapsed_in_sec: u64,
/// The total number of views during benchmarking
pub total_num_views: usize,
/// The number of failed views during benchmarking
pub failed_num_views: usize,
/// The membership committee type used
pub committee_type: String,
}
impl BenchResults {
/// printout the results of one example run
pub fn printout(&self) {
println!("=====================");
println!("{0} Benchmark results:", self.partial_results);
println!("Committee type: {}", self.committee_type);
println!(
"Average latency: {} seconds, Minimum latency: {} seconds, Maximum latency: {} seconds",
self.avg_latency_in_sec, self.minimum_latency_in_sec, self.maximum_latency_in_sec
);
println!("Throughput: {} bytes/sec", self.throughput_bytes_per_sec);
println!(
"Total transactions committed: {}",
self.total_transactions_committed
);
println!(
"Total number of views: {}, Failed number of views: {}",
self.total_num_views, self.failed_num_views
);
println!("=====================");
}
}
/// Struct describing a benchmark result needed for download, also include the config
#[derive(serde::Serialize, serde::Deserialize, Clone, Debug, Default, PartialEq)]
pub struct BenchResultsDownloadConfig {
// Config starting here
/// The commit this benchmark was run on
pub commit_sha: String,
/// Total number of nodes
pub total_nodes: usize,
/// The size of the da committee
pub da_committee_size: usize,
/// The number of fixed_leader_for_gpuvid when we enable the feature [fixed-leader-election]
pub fixed_leader_for_gpuvid: usize,
/// Number of transactions submitted per round
pub transactions_per_round: usize,
/// The size of each transaction in bytes
pub transaction_size: u64,
/// The number of rounds
pub rounds: usize,
// Results starting here
/// Whether the results are partially collected
/// "One" when the results are collected for one node
/// "Half" when the results are collective for half running nodes if not all nodes terminate successfully
/// "Full" if the results are successfully collected from all nodes
pub partial_results: String,
/// The average latency of the transactions
pub avg_latency_in_sec: i64,
/// The minimum latency of the transactions
pub minimum_latency_in_sec: i64,
/// The maximum latency of the transactions
pub maximum_latency_in_sec: i64,
/// The throughput of the consensus protocol = number of transactions committed per second * transaction size in bytes
pub throughput_bytes_per_sec: u64,
/// The number of transactions committed during benchmarking
pub total_transactions_committed: u64,
/// The total time elapsed for benchmarking
pub total_time_elapsed_in_sec: u64,
/// The total number of views during benchmarking
pub total_num_views: usize,
/// The number of failed views during benchmarking
pub failed_num_views: usize,
/// The membership committee type used
pub committee_type: String,
}
// VALIDATOR
#[derive(Parser, Debug, Clone)]
#[command(
name = "Multi-machine consensus",
about = "Simulates consensus among multiple machines"
)]
/// Arguments passed to the validator
pub struct ValidatorArgs {
/// The address the orchestrator runs on
pub url: Url,
/// The optional advertise address to use for Libp2p
pub advertise_address: Option<String>,
/// Optional address to run builder on. Address must be accessible by other nodes
pub builder_address: Option<SocketAddr>,
/// An optional network config file to save to/load from
/// Allows for rejoining the network on a complete state loss
#[arg(short, long)]
pub network_config_file: Option<String>,
}
/// arguments to run multiple validators
#[derive(Parser, Debug, Clone)]
pub struct MultiValidatorArgs {
/// Number of validators to run
pub num_nodes: u16,
/// The address the orchestrator runs on
pub url: Url,
/// The optional advertise address to use for Libp2p
pub advertise_address: Option<String>,
/// An optional network config file to save to/load from
/// Allows for rejoining the network on a complete state loss
#[arg(short, long)]
pub network_config_file: Option<String>,
}
/// Asynchronously retrieves a `NetworkConfig` from an orchestrator.
/// The retrieved one includes correct `node_index` and peer's public config.
///
/// # Errors
/// If we are unable to get the configuration from the orchestrator
pub async fn get_complete_config<K: SignatureKey>(
client: &OrchestratorClient,
mut validator_config: ValidatorConfig<K>,
libp2p_advertise_address: Option<Multiaddr>,
libp2p_public_key: Option<PeerId>,
) -> anyhow::Result<(NetworkConfig<K>, ValidatorConfig<K>, NetworkConfigSource)> {
// get the configuration from the orchestrator
let run_config: NetworkConfig<K> = client
.post_and_wait_all_public_keys::<K>(
&mut validator_config,
libp2p_advertise_address,
libp2p_public_key,
)
.await;
info!(
"Retrieved config; our node index is {}. DA committee member: {}",
run_config.node_index, validator_config.is_da
);
Ok((
run_config,
validator_config,
NetworkConfigSource::Orchestrator,
))
}
impl ValidatorArgs {
/// Constructs `ValidatorArgs` from `MultiValidatorArgs` and a node index.
///
/// If `network_config_file` is present in `MultiValidatorArgs`, it appends the node index to it to create a unique file name for each node.
///
/// # Arguments
///
/// * `multi_args` - A `MultiValidatorArgs` instance containing the base arguments for the construction.
/// * `node_index` - A `u16` representing the index of the node for which the args are being constructed.
///
/// # Returns
///
/// This function returns a new instance of `ValidatorArgs`.
///
/// # Examples
///
/// ```ignore
/// // NOTE this is a toy example,
/// // the user will need to construct a multivalidatorargs since `new` does not exist
/// # use hotshot_orchestrator::client::MultiValidatorArgs;
/// let multi_args = MultiValidatorArgs::new();
/// let node_index = 1;
/// let instance = Self::from_multi_args(multi_args, node_index);
/// ```
#[must_use]
pub fn from_multi_args(multi_args: MultiValidatorArgs, node_index: u16) -> Self {
Self {
url: multi_args.url,
advertise_address: multi_args.advertise_address,
builder_address: None,
network_config_file: multi_args
.network_config_file
.map(|s| format!("{s}-{node_index}")),
}
}
}
impl OrchestratorClient {
/// Creates the client that will connect to the orchestrator
#[must_use]
pub fn new(url: Url) -> Self {
let client = surf_disco::Client::<ClientError, OrchestratorVersion>::new(url);
// TODO ED: Add healthcheck wait here
OrchestratorClient { client }
}
/// Get the config from the orchestrator.
/// If the identity is provided, register the identity with the orchestrator.
/// If not, just retrieving the config (for passive observers)
///
/// # Panics
/// if unable to convert the node index from usize into u64
/// (only applicable on 32 bit systems)
///
/// # Errors
/// If we were unable to serialize the Libp2p data
#[allow(clippy::type_complexity)]
pub async fn get_config_without_peer<K: SignatureKey>(
&self,
libp2p_advertise_address: Option<Multiaddr>,
libp2p_public_key: Option<PeerId>,
) -> anyhow::Result<NetworkConfig<K>> {
// Serialize our (possible) libp2p-specific data
let request_body = vbs::Serializer::<OrchestratorVersion>::serialize(&(
libp2p_advertise_address,
libp2p_public_key,
))?;
let identity = |client: Client<ClientError, OrchestratorVersion>| {
// We need to clone here to move it into the closure
let request_body = request_body.clone();
async move {
let node_index: Result<u16, ClientError> = client
.post("api/identity")
.body_binary(&request_body)
.expect("failed to set request body")
.send()
.await;
node_index
}
.boxed()
};
let node_index = self.wait_for_fn_from_orchestrator(identity).await;
// get the corresponding config
let f = |client: Client<ClientError, OrchestratorVersion>| {
async move {
let config: Result<NetworkConfig<K>, ClientError> = client
.post(&format!("api/config/{node_index}"))
.send()
.await;
config
}
.boxed()
};
let mut config = self.wait_for_fn_from_orchestrator(f).await;
config.node_index = From::<u16>::from(node_index);
Ok(config)
}
/// Post to the orchestrator and get the latest `node_index`
/// Then return it for the init validator config
/// # Panics
/// if unable to post
#[instrument(skip_all, name = "orchestrator node index for validator config")]
pub async fn get_node_index_for_init_validator_config(&self) -> u16 {
let cur_node_index = |client: Client<ClientError, OrchestratorVersion>| {
async move {
let cur_node_index: Result<u16, ClientError> = client
.post("api/get_tmp_node_index")
.send()
.await
.inspect_err(|err| tracing::error!("{err}"));
cur_node_index
}
.boxed()
};
self.wait_for_fn_from_orchestrator(cur_node_index).await
}
/// Requests the configuration from the orchestrator with the stipulation that
/// a successful call requires all nodes to be registered.
///
/// Does not fail, retries internally until success.
#[instrument(skip_all, name = "orchestrator config")]
pub async fn get_config_after_collection<K: SignatureKey>(&self) -> NetworkConfig<K> {
// Define the request for post-register configurations
let get_config_after_collection = |client: Client<ClientError, OrchestratorVersion>| {
async move {
let result = client
.post("api/post_config_after_peer_collected")
.send()
.await;
if let Err(ref err) = result {
tracing::error!("{err}");
}
result
}
.boxed()
};
// Loop until successful
self.wait_for_fn_from_orchestrator(get_config_after_collection)
.await
}
/// Registers a builder URL with the orchestrator
///
/// # Panics
/// if unable to serialize `address`
pub async fn post_builder_addresses(&self, addresses: Vec<Url>) {
let send_builder_f = |client: Client<ClientError, OrchestratorVersion>| {
let request_body = vbs::Serializer::<OrchestratorVersion>::serialize(&addresses)
.expect("Failed to serialize request");
async move {
let result: Result<_, ClientError> = client
.post("api/builder")
.body_binary(&request_body)
.unwrap()
.send()
.await
.inspect_err(|err| tracing::error!("{err}"));
result
}
.boxed()
};
self.wait_for_fn_from_orchestrator::<_, _, ()>(send_builder_f)
.await;
}
/// Requests a builder URL from orchestrator
pub async fn get_builder_addresses(&self) -> Vec<Url> {
// Define the request for post-register configurations
let get_builder = |client: Client<ClientError, OrchestratorVersion>| {
async move {
let result = client.get("api/builders").send().await;
if let Err(ref err) = result {
tracing::error!("{err}");
}
result
}
.boxed()
};
// Loop until successful
self.wait_for_fn_from_orchestrator(get_builder).await
}
/// Sends my public key to the orchestrator so that it can collect all public keys
/// And get the updated config
/// Blocks until the orchestrator collects all peer's public keys/configs
/// # Panics
/// if unable to post
#[instrument(skip(self), name = "orchestrator public keys")]
pub async fn post_and_wait_all_public_keys<K: SignatureKey>(
&self,
validator_config: &mut ValidatorConfig<K>,
libp2p_advertise_address: Option<Multiaddr>,
libp2p_public_key: Option<PeerId>,
) -> NetworkConfig<K> {
let pubkey: Vec<u8> = PeerConfig::<K>::to_bytes(&validator_config.public_config()).clone();
let da_requested: bool = validator_config.is_da;
// Serialize our (possible) libp2p-specific data
let request_body = vbs::Serializer::<OrchestratorVersion>::serialize(&(
pubkey,
libp2p_advertise_address,
libp2p_public_key,
))
.expect("failed to serialize request");
// register our public key with the orchestrator
let (node_index, is_da): (u64, bool) = loop {
let result = self
.client
.post(&format!("api/pubkey/{da_requested}"))
.body_binary(&request_body)
.expect("Failed to form request")
.send()
.await
.inspect_err(|err| tracing::error!("{err}"));
if let Ok((index, is_da)) = result {
break (index, is_da);
}
sleep(Duration::from_millis(250)).await;
};
validator_config.is_da = is_da;
// wait for all nodes' public keys
let wait_for_all_nodes_pub_key = |client: Client<ClientError, OrchestratorVersion>| {
async move {
client
.get("api/peer_pub_ready")
.send()
.await
.inspect_err(|err| tracing::error!("{err}"))
}
.boxed()
};
self.wait_for_fn_from_orchestrator::<_, _, ()>(wait_for_all_nodes_pub_key)
.await;
let mut network_config = self.get_config_after_collection().await;
network_config.node_index = node_index;
network_config
}
/// Tells the orchestrator this validator is ready to start
/// Blocks until the orchestrator indicates all nodes are ready to start
/// # Panics
/// Panics if unable to post.
#[instrument(skip(self), name = "orchestrator ready signal")]
pub async fn wait_for_all_nodes_ready(&self, peer_config: Vec<u8>) -> bool {
let send_ready_f = |client: Client<ClientError, OrchestratorVersion>| {
let pk = peer_config.clone();
async move {
let result: Result<_, ClientError> = client
.post("api/ready")
.body_binary(&pk)
.unwrap()
.send()
.await
.inspect_err(|err| tracing::error!("{err}"));
result
}
.boxed()
};
self.wait_for_fn_from_orchestrator::<_, _, ()>(send_ready_f)
.await;
let wait_for_all_nodes_ready_f = |client: Client<ClientError, OrchestratorVersion>| {
async move { client.get("api/start").send().await }.boxed()
};
self.wait_for_fn_from_orchestrator(wait_for_all_nodes_ready_f)
.await
}
/// Sends the benchmark metrics to the orchestrator
/// # Panics
/// Panics if unable to post
#[instrument(skip_all, name = "orchestrator metrics")]
pub async fn post_bench_results(&self, bench_results: BenchResults) {
let _send_metrics_f: Result<(), ClientError> = self
.client
.post("api/results")
.body_json(&bench_results)
.unwrap()
.send()
.await
.inspect_err(|err| tracing::warn!("{err}"));
}
/// Generic function that waits for the orchestrator to return a non-error
/// Returns whatever type the given function returns
#[instrument(skip_all, name = "waiting for orchestrator")]
async fn wait_for_fn_from_orchestrator<F, Fut, GEN>(&self, f: F) -> GEN
where
F: Fn(Client<ClientError, OrchestratorVersion>) -> Fut,
Fut: Future<Output = Result<GEN, ClientError>>,
{
loop {
let client = self.client.clone();
let res = f(client).await;
match res {
Ok(x) => break x,
Err(err) => {
tracing::info!("{err}");
sleep(Duration::from_millis(250)).await;
}
}
}
}
}