refactor appState and proxy

Cargo.toml

@@ -31,6 +31,9 @@ serde = "1.0.219"
 serde_json = "1.0"
 sqlparser = "0.51"
 sqlx = { version = "0.8", features = ["runtime-tokio-rustls", "postgres", "uuid", "time"] }
+tracing = "0.1"
+tracing-log = "0.1"
+tracing-subscriber = "0.3.20"
 thiserror = "2.0.12"
 tiktoken-rs = "0.7.0"
 tokio = { version = "1.0", features = ["full"] }

extension/Makefile

@@ -42,6 +42,7 @@ clean:
 setup.dependencies: install-pg_cron install-pgvector install-pgmq install-vectorscale
 setup.shared_preload_libraries:
 	echo "shared_preload_libraries = 'pg_cron, vectorize'" >> ~/.pgrx/data-${PG_VERSION}/postgresql.conf
+	echo "cron.database_name = 'postgres'" >> ~/.pgrx/data-${PG_VERSION}/postgresql.conf
 setup.urls:
 	echo "vectorize.embedding_service_url = 'http://localhost:3000/v1'" >> ~/.pgrx/data-${PG_VERSION}/postgresql.conf
 	echo "vectorize.ollama_service_url = 'http://localhost:3001'" >> ~/.pgrx/data-${PG_VERSION}/postgresql.conf
@@ -93,7 +94,7 @@ test-integration:
 	cargo test ${TEST_NAME} -- --ignored --test-threads=1 --nocapture
 
 test-unit:
-	cargo test ${TEST_NAME} -- --test-threads=1
+	cargo test ${TEST_NAME} -- --test-threads=1 --nocapture
 
 test-version:
 	git fetch --tags

extension/src/api.rs

@@ -176,40 +176,6 @@ fn encode(
     Ok(transform(input, &model, api_key).remove(0))
 }
 
-#[allow(clippy::too_many_arguments)]
-#[deprecated(since = "0.22.0", note = "Please use vectorize.table() instead")]
-#[pg_extern]
-fn init_rag(
-    agent_name: &str,
-    table_name: &str,
-    unique_record_id: &str,
-    // column that have data we want to be able to chat with
-    column: &str,
-    schema: default!(&str, "'public'"),
-    index_dist_type: default!(types::IndexDist, "'pgv_hnsw_cosine'"),
-    // transformer model to use in vector-search
-    transformer: default!(&str, "'sentence-transformers/all-MiniLM-L6-v2'"),
-    table_method: default!(types::TableMethod, "'join'"),
-    schedule: default!(&str, "'* * * * *'"),
-) -> Result<String> {
-    pgrx::warning!("DEPRECATED: vectorize.init_rag() will be removed in a future version. Please use vectorize.table() instead.");
-    // chat only supports single columns transform
-    let columns = vec![column.to_string()];
-    let transformer_model = Model::new(transformer)?;
-    init_table(
-        agent_name,
-        schema,
-        table_name,
-        columns,
-        unique_record_id,
-        None,
-        index_dist_type.into(),
-        &transformer_model,
-        table_method.into(),
-        schedule,
-    )
-}
-
 /// creates a table indexed with embeddings for chat completion workloads
 #[pg_extern]
 fn rag(

extension/src/executor.rs

@@ -25,7 +25,7 @@ pub fn batch_texts(
         return TableIterator::new(vec![record_ids].into_iter().map(|arr| (arr,)));
     }
 
-    let num_batches = (total_records + batch_size - 1) / batch_size;
+    let num_batches = total_records.div_ceil(batch_size);
 
     let mut batches = Vec::with_capacity(num_batches);
 

extension/src/guc.rs

@@ -250,7 +250,6 @@ pub fn get_guc(guc: VectorizeGuc) -> Option<String> {
     }
 }
 
-#[allow(dead_code)]
 fn handle_cstr(cstr: &CStr) -> Result<String> {
     if let Ok(s) = cstr.to_str() {
         Ok(s.to_owned())

extension/src/search.rs

@@ -429,7 +429,7 @@ pub fn cosine_similarity_search(
             num_results,
             where_clause,
         ),
-        TableMethod::join => query::join_table_cosine_similarity(
+        TableMethod::join => join_table_cosine_similarity(
             project,
             &job_params.schema,
             &job_params.relation,
@@ -452,6 +452,52 @@ pub fn cosine_similarity_search(
     })
 }
 
+pub fn join_table_cosine_similarity(
+    project: &str,
+    schema: &str,
+    table: &str,
+    join_key: &str,
+    return_columns: &[String],
+    num_results: i32,
+    where_clause: Option<String>,
+) -> String {
+    let cols = &return_columns
+        .iter()
+        .map(|s| format!("t0.{s}"))
+        .collect::<Vec<_>>()
+        .join(",");
+    let where_str = if let Some(w) = where_clause {
+        prepare_filter(&w, join_key)
+    } else {
+        "".to_string()
+    };
+    let inner_query = format!(
+        "
+    SELECT
+        {join_key},
+        1 - (embeddings <=> $1::vector) AS similarity_score
+    FROM vectorize._embeddings_{project}
+    ORDER BY similarity_score DESC
+    "
+    );
+    format!(
+        "
+    SELECT to_jsonb(t) as results
+    FROM (
+        SELECT {cols}, t1.similarity_score
+        FROM
+            (
+                {inner_query}
+            ) t1
+        INNER JOIN {schema}.{table} t0 on t0.{join_key} = t1.{join_key}
+        {where_str}
+    ) t
+    ORDER BY t.similarity_score DESC
+    LIMIT {num_results};
+    "
+    )
+}
+
 fn single_table_cosine_similarity(
     project: &str,
     schema: &str,
@@ -482,3 +528,9 @@ fn single_table_cosine_similarity(
         cols = return_columns.join(", "),
     )
 }
+
+// transform user's where_sql into the format search query expects
+fn prepare_filter(filter: &str, pkey: &str) -> String {
+    let wc = filter.replace(pkey, &format!("t0.{pkey}"));
+    format!("AND {wc}")
+}

extension/tests/util.rs

@@ -7,7 +7,6 @@ pub mod common {
     use sqlx::{Pool, Postgres, Row};
     use url::{ParseError, Url};
 
-    #[allow(dead_code)]
     #[derive(FromRow, Debug, serde::Deserialize)]
     pub struct SearchResult {
         pub product_id: i32,
@@ -16,7 +15,6 @@ pub mod common {
         pub similarity_score: f64,
     }
 
-    #[allow(dead_code)]
     #[derive(FromRow, Debug, Serialize)]
     pub struct SearchJSON {
         pub search_results: serde_json::Value,

proxy/Cargo.toml

@@ -13,5 +13,8 @@ serde_json = { workspace = true }
 sqlx = { workspace = true}
 thiserror = { workspace = true }
 tokio = { workspace = true }
+tracing = { workspace = true }
+tracing-subscriber = { workspace = true }
+url = { workspace = true }
 
 pgwire = { version = "0.30", features = ["server-api-aws-lc-rs"] }

proxy/src/proxy.rs

@@ -1,8 +1,15 @@
-use log::{error, info};
+use std::collections::HashMap;
+use std::net::SocketAddr;
+use std::net::ToSocketAddrs;
 use std::sync::Arc;
+use std::time::Duration;
 use tokio::io::{AsyncRead, AsyncReadExt, AsyncWrite, AsyncWriteExt};
-use tokio::net::TcpStream;
+use tokio::net::{TcpListener, TcpStream};
+use tokio::sync::RwLock;
 use tokio::time::timeout;
+use tracing::{error, info};
+use url::Url;
+use vectorize_core::types::VectorizeJob;
 
 use super::message_parser::{log_message_processing, try_parse_complete_message};
 use super::protocol::{BUFFER_SIZE, ProxyConfig, WireProxyError};
@@ -129,3 +136,55 @@ where
     info!("Standard proxy stream closed: {total_bytes} bytes transferred");
     Ok(())
 }
+
+pub async fn start_postgres_proxy(
+    proxy_port: u16,
+    database_url: String,
+    job_cache: Arc<RwLock<HashMap<String, VectorizeJob>>>,
+    db_pool: sqlx::PgPool,
+) -> Result<(), Box<dyn std::error::Error>> {
+    let bind_address = "0.0.0.0";
+    let timeout = 30;
+
+    let listen_addr: SocketAddr = format!("{}:{}", bind_address, proxy_port).parse()?;
+
+    let url = Url::parse(&database_url)?;
+    let postgres_host = url.host_str().unwrap();
+    let postgres_port = url.port().unwrap();
+
+    let postgres_addr: SocketAddr = format!("{postgres_host}:{postgres_port}")
+        .to_socket_addrs()?
+        .next()
+        .ok_or("Failed to resolve PostgreSQL host address")?;
+
+    let config = Arc::new(ProxyConfig {
+        postgres_addr,
+        timeout: Duration::from_secs(timeout),
+        jobmap: job_cache,
+        db_pool,
+        prepared_statements: Arc::new(RwLock::new(HashMap::new())),
+    });
+
+    info!("Proxy listening on: {listen_addr}");
+    info!("Forwarding to PostgreSQL at: {postgres_addr}");
+
+    let listener = TcpListener::bind(listen_addr).await?;
+
+    loop {
+        match listener.accept().await {
+            Ok((client_stream, client_addr)) => {
+                info!("New proxy connection from: {client_addr}");
+
+                let config = Arc::clone(&config);
+                tokio::spawn(async move {
+                    if let Err(e) = handle_connection_with_timeout(client_stream, config).await {
+                        error!("Proxy connection error from {client_addr}: {e}");
+                    }
+                });
+            }
+            Err(e) => {
+                error!("Failed to accept proxy connection: {e}");
+            }
+        }
+    }
+}

server/Cargo.toml

@@ -14,6 +14,9 @@ vectorize_core = { package = "vectorize-core", path = "../core" }
 vectorize_worker = { package = "vectorize-worker", path = "../worker" }
 vectorize_proxy = { package = "vectorize-proxy", path = "../proxy" }
 pgmq = { workspace = true }
+tracing-subscriber = { workspace = true }
+tracing = { workspace = true }
+tracing-log = { workspace = true }
 
 actix-cors = "0.7.1"
 actix-http = "3.11.0"
@@ -26,8 +29,6 @@ bytes = "1.10.1"
 chrono = {version = "0.4.41", features = ["serde"] }
 clap = { version = "4.0", features = ["derive"] }
 env = "1.0.1"
-tracing = "0.1"
-tracing-log = "0.1"
 fallible-iterator = "0.3.0"
 futures = "0.3.31"
 lazy_static = "1.5.0"
@@ -46,7 +47,6 @@ thiserror = "2.0.12"
 tiktoken-rs = "0.7.0"
 tokio = { version = "1.0", features = ["full"] }
 tokio-postgres = "0.7"
-tracing-subscriber = "0.3"
 url = "2.2"
 utoipa = { version = "4", features = ["actix_extras", "chrono", "uuid"] }
 utoipa-swagger-ui  = { version = "7", features = ["actix-web"] }