feature(rpc): Migrate from deprecated jsonrpc_* crates to `jsonrpse…

…e` (#9059)

* update methods

* update get block template rpc methods

* update other getblocktemplate files

* upgrade server and middlewares

* upgrade zebrad start command

* remove unused imports

* add a todo for unauthenticated rpc error

* upgrade tests, temporally comment out some of them

* fix the rpc tx queue

* update denies

* fix  links

* clippy

* fir more doc links

* fix queue tests

Co-authored-by: Arya <aryasolhi@gmail.com>

* add suggestions from code review

* fix snapshots

* try `block_on` instead of `now_or_never` in the http middleware

* move import

* Apply suggestions from code review

Co-authored-by: Arya <aryasolhi@gmail.com>

* fix bounds

---------

Co-authored-by: Arya <aryasolhi@gmail.com>

deny.toml

@@ -78,19 +78,8 @@ skip-tree = [
     { name = "base64", version = "=0.21.7" },
     { name = "sync_wrapper", version = "0.1.2" },
 
-    # wait for jsonrpc-http-server to update hyper or for Zebra to replace jsonrpc (#8682)
-    { name = "h2", version = "=0.3.26" },
-    { name = "http", version = "=0.2.12" },
-    { name = "http-body", version = "=0.4.6" },
-    { name = "hyper", version = "=0.14.31" },
-    { name = "hyper-rustls", version = "=0.24.2" },
-
-    { name = "reqwest", version = "=0.11.27" },
-    { name = "rustls", version = "=0.21.12" },
-    { name = "rustls-pemfile", version = "=1.0.4" },
-    { name = "rustls-webpki", version = "=0.101.7" },
-    { name = "tokio-rustls", version = "=0.24.1" },
-    { name = "webpki-roots", version = "=0.25.4" },
+    # wait for abscissa_core to update toml
+    { name = "toml", version = "=0.5.11" },
 
     # wait for structopt-derive to update heck
     { name = "heck", version = "=0.3.3" },

zebra-rpc/Cargo.toml

@@ -59,9 +59,11 @@ chrono = { version = "0.4.39", default-features = false, features = [
 ] }
 futures = "0.3.31"
 
-jsonrpc-core = "18.0.0"
-jsonrpc-derive = "18.0.0"
-jsonrpc-http-server = "18.0.0"
+jsonrpsee = { version = "0.24.7", features = ["server"] }
+jsonrpsee-types = "0.24.7"
+jsonrpsee-proc-macros = "0.24.7"
+hyper = "1.5.0"
+http-body-util = "0.1.2"
 
 # zebra-rpc needs the preserve_order feature in serde_json, which is a dependency of jsonrpc-core
 serde_json = { version = "1.0.133", features = ["preserve_order"] }

zebra-rpc/src/config.rs

@@ -50,24 +50,12 @@ pub struct Config {
 
     /// The number of threads used to process RPC requests and responses.
     ///
-    /// Zebra's RPC server has a separate thread pool and a `tokio` executor for each thread.
-    /// State queries are run concurrently using the shared thread pool controlled by
-    /// the [`SyncSection.parallel_cpu_threads`](https://docs.rs/zebrad/latest/zebrad/components/sync/struct.Config.html#structfield.parallel_cpu_threads) config.
-    ///
-    /// If the number of threads is not configured or zero, Zebra uses the number of logical cores.
-    /// If the number of logical cores can't be detected, Zebra uses one thread.
-    ///
-    /// Set to `1` to run all RPC queries on a single thread, and detect RPC port conflicts from
-    /// multiple Zebra or `zcashd` instances.
-    ///
-    /// For details, see [the `jsonrpc_http_server` documentation](https://docs.rs/jsonrpc-http-server/latest/jsonrpc_http_server/struct.ServerBuilder.html#method.threads).
-    ///
-    /// ## Warning
-    ///
-    /// The default config uses multiple threads, which disables RPC port conflict detection.
-    /// This can allow multiple Zebra instances to share the same RPC port.
-    ///
-    /// If some of those instances are outdated or failed, RPC queries can be slow or inconsistent.
+    /// This field is deprecated and could be removed in a future release.
+    /// We keep it just for backward compatibility but it actually do nothing.
+    /// It was something configurable when the RPC server was based in the jsonrpc-core crate,
+    /// not anymore since we migrated to jsonrpsee.
+    // TODO: Prefix this field name with an underscore so it's clear that it's now unused, and
+    //       use serde(rename) to continue successfully deserializing old configs.
     pub parallel_cpu_threads: usize,
 
     /// Test-only option that makes Zebra say it is at the chain tip,