feat(avm)!: improve anomaly detection (#479)

Cargo.lock

@@ -65,7 +65,7 @@ version = "0.1.1"
 dependencies = [
  "air-beautifier",
  "anyhow",
- "clap 4.1.4",
+ "clap 4.1.6",
 ]
 
 [[package]]
@@ -233,7 +233,7 @@ dependencies = [
  "anyhow",
  "avm-data-store",
  "avm-interface",
- "clap 4.1.4",
+ "clap 4.1.6",
  "itertools",
  "serde",
  "serde_json",
@@ -333,6 +333,7 @@ checksum = "d468802bab17cbc0cc575e9b053f41e72aa36bfa6b7f55e3529ffa43161b97fa"
 name = "avm-data-store"
 version = "0.4.1"
 dependencies = [
+ "avm-interface",
  "serde",
  "serde_bytes",
  "serde_json",
@@ -344,7 +345,6 @@ version = "0.28.1"
 dependencies = [
  "air-interpreter-interface",
  "air-utils",
- "avm-data-store",
  "log",
  "maplit",
  "polyplets",
@@ -452,18 +452,6 @@ version = "2.4.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "cfa8873f51c92e232f9bac4065cddef41b714152812bfc5f7672ba16d6ef8cd9"
 
-[[package]]
-name = "bstr"
-version = "0.2.17"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ba3569f383e8f1598449f1a423e72e99569137b47740b1da11ef19af3d5c3223"
-dependencies = [
- "lazy_static",
- "memchr",
- "regex-automata",
- "serde",
-]
-
 [[package]]
 name = "bumpalo"
 version = "3.12.0"
@@ -596,9 +584,9 @@ dependencies = [
 
 [[package]]
 name = "clap"
-version = "4.1.4"
+version = "4.1.6"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f13b9c79b5d1dd500d20ef541215a6423c75829ef43117e1b4d17fd8af0b5d76"
+checksum = "ec0b0588d44d4d63a87dbd75c136c166bbfd9a86a31cb89e09906521c7d3f5e3"
 dependencies = [
  "bitflags",
  "clap_derive",
@@ -911,13 +899,12 @@ dependencies = [
 
 [[package]]
 name = "csv"
-version = "1.1.6"
+version = "1.2.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "22813a6dc45b335f9bade10bf7271dc477e81113e89eb251a0bc2a8a81c536e1"
+checksum = "af91f40b7355f82b0a891f50e70399475945bb0b0da4f1700ce60761c9d3e359"
 dependencies = [
- "bstr",
  "csv-core",
- "itoa 0.4.8",
+ "itoa",
  "ryu",
  "serde",
 ]
@@ -943,9 +930,9 @@ dependencies = [
 
 [[package]]
 name = "cxx"
-version = "1.0.89"
+version = "1.0.91"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "bc831ee6a32dd495436e317595e639a587aa9907bef96fe6e6abc290ab6204e9"
+checksum = "86d3488e7665a7a483b57e25bdd90d0aeb2bc7608c8d0346acf2ad3f1caf1d62"
 dependencies = [
  "cc",
  "cxxbridge-flags",
@@ -955,9 +942,9 @@ dependencies = [
 
 [[package]]
 name = "cxx-build"
-version = "1.0.89"
+version = "1.0.91"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "94331d54f1b1a8895cd81049f7eaaaef9d05a7dcb4d1fd08bf3ff0806246789d"
+checksum = "48fcaf066a053a41a81dfb14d57d99738b767febb8b735c3016e469fac5da690"
 dependencies = [
  "cc",
  "codespan-reporting",
@@ -970,15 +957,15 @@ dependencies = [
 
 [[package]]
 name = "cxxbridge-flags"
-version = "1.0.89"
+version = "1.0.91"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "48dcd35ba14ca9b40d6e4b4b39961f23d835dbb8eed74565ded361d93e1feb8a"
+checksum = "a2ef98b8b717a829ca5603af80e1f9e2e48013ab227b68ef37872ef84ee479bf"
 
 [[package]]
 name = "cxxbridge-macro"
-version = "1.0.89"
+version = "1.0.91"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "81bbeb29798b407ccd82a3324ade1a7286e0d29851475990b612670f6f5124d2"
+checksum = "086c685979a698443656e5cf7856c95c642295a38599f12fb1ff76fb28d19892"
 dependencies = [
  "proc-macro2",
  "quote",
@@ -1183,9 +1170,9 @@ checksum = "0ce7134b9999ecaf8bcd65542e436736ef32ddca1b3e06094cb6ec5755203b80"
 
 [[package]]
 name = "fluence-app-service"
-version = "0.23.0"
+version = "0.23.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "3264bb11b79b9e3e4e5b722d986efd461e4606477ba2e454e61ddf79bc1d3bf1"
+checksum = "942caba094d1c219ad3656d51c827470d12dac834b4325843e9f61b12bf0d0dd"
 dependencies = [
  "log",
  "maplit",
@@ -1521,12 +1508,6 @@ dependencies = [
  "either",
 ]
 
-[[package]]
-name = "itoa"
-version = "0.4.8"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b71991ff56294aa922b450139ee08b3bfc70982c6b2c7562771375cf73542dd4"
-
 [[package]]
 name = "itoa"
 version = "1.0.5"
@@ -1824,9 +1805,9 @@ dependencies = [
 
 [[package]]
 name = "marine-runtime"
-version = "0.24.0"
+version = "0.24.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ba3726e9dccce778d0b848ebc9adb312f33f338128bb182aa5ed8429f3f7aac1"
+checksum = "545262bf5b68d2b4f4b9eed2557f09421cc4bb62a10e71c307140f27b6411b9c"
 dependencies = [
  "bytesize",
  "it-json-serde",
@@ -2053,9 +2034,9 @@ dependencies = [
 
 [[package]]
 name = "once_cell"
-version = "1.17.0"
+version = "1.17.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "6f61fba1741ea2b3d6a1e3178721804bb716a68a6aeba1149b5d52e3d464ea66"
+checksum = "b7e5500299e16ebb147ae15a00a942af264cf3688f47923b8fc2cd5858f23ad3"
 
 [[package]]
 name = "oorandom"
@@ -2577,7 +2558,7 @@ version = "1.0.93"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "cad406b69c91885b5107daf2c29572f6c8cdb3c66826821e286c533490c0bc76"
 dependencies = [
- "itoa 1.0.5",
+ "itoa",
  "ryu",
  "serde",
 ]
@@ -2595,7 +2576,7 @@ dependencies = [
  "serde",
  "serde_json",
  "serde_with_macros",
- "time 0.3.17",
+ "time 0.3.19",
 ]
 
 [[package]]
@@ -2781,10 +2762,11 @@ dependencies = [
 
 [[package]]
 name = "thread_local"
-version = "1.1.4"
+version = "1.1.7"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5516c27b78311c50bf42c071425c560ac799b11c30b31f87e3081965fe5e0180"
+checksum = "3fdd6f064ccff2d6567adcb3873ca630700f00b5ad3f060c25b5dcfd9a4ce152"
 dependencies = [
+ "cfg-if 1.0.0",
  "once_cell",
 ]
 
@@ -2801,11 +2783,11 @@ dependencies = [
 
 [[package]]
 name = "time"
-version = "0.3.17"
+version = "0.3.19"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "a561bf4617eebd33bca6434b988f39ed798e527f51a1e797d0ee4f61c0a38376"
+checksum = "53250a3b3fed8ff8fd988587d8925d26a83ac3845d9e03b220b37f34c2b8d6c2"
 dependencies = [
- "itoa 1.0.5",
+ "itoa",
  "serde",
  "time-core",
  "time-macros",
@@ -2819,9 +2801,9 @@ checksum = "2e153e1f1acaef8acc537e68b44906d2db6436e2b35ac2c6b42640fff91f00fd"
 
 [[package]]
 name = "time-macros"
-version = "0.2.6"
+version = "0.2.7"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d967f99f534ca7e495c575c62638eebc2898a8c84c119b89e250477bc4ba16b2"
+checksum = "a460aeb8de6dcb0f381e1ee05f1cd56fcf5a5f6eb8187ff3d8f0b11078d38b7c"
 dependencies = [
  "time-core",
 ]
@@ -2911,7 +2893,7 @@ dependencies = [
  "sharded-slab",
  "smallvec",
  "thread_local",
- "time 0.3.17",
+ "time 0.3.19",
  "tracing",
  "tracing-core",
  "tracing-serde",

avm/interface/Cargo.toml

@@ -17,7 +17,6 @@ path = "src/lib.rs"
 [dependencies]
 air-interpreter-interface = { version = "0.12.1", path = "../../crates/air-lib/interpreter-interface", default-features = false }
 air-utils = { version = "0.1.0", path = "../../crates/air-lib/utils" }
-avm-data-store = { version = "0.4.1", path = "../../crates/data-store" }
 polyplets = { version = "0.3.2", path = "../../crates/air-lib/polyplets" }
 
 thiserror = "1.0.38"

avm/server/src/avm.rs

@@ -102,16 +102,20 @@ impl<E> AVM<E> {
                 particle_parameters.timestamp,
                 particle_parameters.ttl,
                 particle_parameters.current_peer_id.clone(),
-                call_results,
+                call_results.clone(),
             )
             .map_err(AVMError::RunnerError)?;
 
         let execution_time = execution_start_time.elapsed();
         let memory_delta = self.memory_stats().memory_size - memory_size_before;
-        if self.data_store.detect_anomaly(execution_time, memory_delta) {
+        if self
+            .data_store
+            .detect_anomaly(execution_time, memory_delta, &outcome)
+        {
             self.save_anomaly_data(
                 &air,
                 &current_data,
+                &call_results,
                 &particle_parameters,
                 &outcome,
                 execution_time,
@@ -139,11 +143,12 @@ impl<E> AVM<E> {
         self.runner.memory_stats()
     }
 
-    #[allow(clippy::result_large_err)]
+    #[allow(clippy::result_large_err, clippy::too_many_arguments)]
     fn save_anomaly_data(
         &mut self,
         air_script: &str,
         current_data: &[u8],
+        call_result: &CallResults,
         particle_parameters: &ParticleParameters<'_>,
         avm_outcome: &RawAVMOutcome,
         execution_time: Duration,
@@ -152,6 +157,7 @@ impl<E> AVM<E> {
         let prev_data = self
             .data_store
             .read_data(&particle_parameters.particle_id)?;
+        let call_results = serde_json::to_vec(call_result).map_err(AVMError::AnomalyDataSeError)?;
         let ser_particle =
             serde_json::to_vec(particle_parameters).map_err(AVMError::AnomalyDataSeError)?;
         let ser_avm_outcome =
@@ -162,6 +168,7 @@ impl<E> AVM<E> {
             &ser_particle,
             &prev_data,
             current_data,
+            &call_results,
             &ser_avm_outcome,
             execution_time,
             memory_delta,

crates/air-lib/air-parser/Cargo.toml

@@ -14,7 +14,7 @@ lalrpop = "0.19.8"
 
 [dependencies]
 air-lambda-ast = { version = "0.1.0", path = "../lambda/ast" }
-air-lambda-parser = { vesrion = "0.1.0", path = "../lambda/parser" }
+air-lambda-parser = { version = "0.1.0", path = "../lambda/parser" }
 
 lalrpop-util = "0.19.8"
 regex = "1.7.1"

crates/data-store/Cargo.toml

@@ -15,6 +15,7 @@ name = "avm_data_store"
 path = "src/lib.rs"
 
 [dependencies]
+avm-interface = { version = "0.28.1", path = "../../avm/interface"}
 serde = { version = "1.0.152", features = ["derive"] }
 serde_bytes = "0.11.9"
 

crates/data-store/src/lib.rs

@@ -26,6 +26,8 @@
     unreachable_patterns
 )]
 
+use avm_interface::raw_outcome::RawAVMOutcome;
+
 use serde::Deserialize;
 use serde::Serialize;
 use std::borrow::Cow;
@@ -48,10 +50,16 @@ pub trait DataStore {
 
     /// Returns true if an anomaly happened and it's necessary to save execution data
     /// for debugging purposes.
-    ///  execution_time - is time taken by the interpreter to execute provided script
+    ///  execution_time - time taken by the interpreter to execute provided script
-    ///  memory_delta - is a count of bytes on which an interpreter heap has been extended
+    ///  memory_delta - count of bytes on which an interpreter heap has been extended
     ///                 during execution of a particle
-    fn detect_anomaly(&self, execution_time: Duration, memory_delta: usize) -> bool;
+    ///  outcome - a result of AquaVM invocation
+    fn detect_anomaly(
+        &self,
+        execution_time: Duration,
+        memory_delta: usize,
+        outcome: &RawAVMOutcome,
+    ) -> bool;
 
     fn collect_anomaly_data(
         &mut self,
@@ -71,17 +79,21 @@ pub struct AnomalyData<'data> {
     #[serde(borrow, with = "serde_bytes")]
     pub current_data: Cow<'data, [u8]>,
     #[serde(borrow, with = "serde_bytes")]
-    pub avm_outcome: Cow<'data, [u8]>, // it's byte because of the restriction on trait objects methods
+    pub call_results: Cow<'data, [u8]>,
+    #[serde(borrow, with = "serde_bytes")]
+    pub avm_outcome: Cow<'data, [u8]>,
     pub execution_time: Duration,
     pub memory_delta: usize,
 }
 
 impl<'data> AnomalyData<'data> {
+    #[allow(clippy::too_many_arguments)]
     pub fn new(
         air_script: &'data str,
         particle: &'data [u8],
         prev_data: &'data [u8],
         current_data: &'data [u8],
+        call_results: &'data [u8],
         avm_outcome: &'data [u8],
         execution_time: Duration,
         memory_delta: usize,
@@ -91,6 +103,7 @@ impl<'data> AnomalyData<'data> {
             particle: particle.into(),
             prev_data: prev_data.into(),
             current_data: current_data.into(),
+            call_results: call_results.into(),
             avm_outcome: avm_outcome.into(),
             execution_time,
             memory_delta,
@@ -107,6 +120,7 @@ mod tests {
         particle: &[u8],
         prev_data: &[u8],
         current_data: &[u8],
+        call_results: &[u8],
         avm_outcome: &[u8],
     ) -> String {
         format!(
@@ -115,6 +129,7 @@ mod tests {
                 r#""particle":{particle},"#,
                 r#""prev_data":{prev_data},"#,
                 r#""current_data":{current_data},"#,
+                r#""call_results":{call_results},"#,
                 r#""avm_outcome":{avm_outcome},"#,
                 r#""execution_time":{{"secs":42,"nanos":0}},"#,
                 r#""memory_delta":123"#,
@@ -124,6 +139,7 @@ mod tests {
             particle = serde_json::to_string(particle).unwrap(),
             prev_data = serde_json::to_string(prev_data).unwrap(),
             current_data = serde_json::to_string(current_data).unwrap(),
+            call_results = serde_json::to_string(call_results).unwrap(),
             avm_outcome = serde_json::to_string(avm_outcome).unwrap(),
         )
     }
@@ -135,6 +151,7 @@ mod tests {
             br#"{"trace":[]}"#,      // not real data
             br#"{"trace":[1,2,3]}"#, // not real data
             b"{}",                   // not real data
+            b"{}",
             Duration::from_secs(42),
             123,
         );
@@ -145,6 +162,7 @@ mod tests {
             &anomaly.particle,
             &anomaly.prev_data,
             &anomaly.current_data,
+            &anomaly.call_results,
             &anomaly.avm_outcome,
         );
         assert_eq!(json_data, expected);
@@ -156,11 +174,13 @@ mod tests {
         let current_data = br#"{"data":"current"}"#;
         let prev_data = br#"{"data":"prev"}"#;
         let avm_outcome = br#"{"avm":[1,2,3]}"#;
+        let call_results = br#"{"call_results": "excellent result"}"#;
         let json_data = anomaly_json(
             "(null)",
             &particle[..],
             &prev_data[..],
             &current_data[..],
+            &call_results[..],
             &avm_outcome[..],
         );
 
@@ -174,6 +194,7 @@ mod tests {
                 &particle[..],
                 &prev_data[..],
                 &current_data[..],
+                &call_results[..],
                 &avm_outcome[..],
                 Duration::from_secs(42),
                 123,