refactor: Upgraded all the dependencies

[frol]

The main goal is to get the latest actix/actix-web along with tokio 1.x. Fixes #3868

Extra bonuses:

1. Resolved compilation errors due to use of private modules of syn and serde crates in our code-base
2. Hopefully it will address a long-standing Telemetry issue "not an error"
3. It may also address a memory leak in Telemetry

[frol]

Nayduck link: http://nayduck.eastus.cloudapp.azure.com:3000/#/run/1064

[birchmd]

Amazing! 🥇

Of note: fixes #3868

[birchmd]

Depending on a random branch seems brittle to me. What happens if the user deletes the branch? Further down you list the dependency as 0.5.4; is there something in particular we need here that is not needed there? If so, maybe we should make our own fork of actix-extras where we can keep this branch around until we are able to upgrade to the next release of actix-cors.

[frol]

I need beta versions of actix with tokio 1.x support, and actix-cors has not been released with the version bump yet. Forking to our guthub org makes perfect sense, I will do that 👍

[frol]

Extra bonuses:

1. Resolved compilation errors due to use of private modules of syn and serde crates in our code-base
2. Hopefully it will address a long-standing Telemetry issue "not an error"
3. It may also address a memory leak in Telemetry

[bowenwang1996]

@frol looks like the PR doesn't yet compile. Are you going to address it any time soon? We need it asap to address #3854

EDIT: looks like neard compiles fine, but it is still desirable to get this PR to a ready state sooner than later

[frol]

It passed initial checks locally, but I expected it to fail somewhere anyway. I left my home office for a walk, and I will be back in 4-5 hours. If you have a bandwidth to work on it before I get back, go for it, otherwise, I was going to address CI issues during the weekend to minimize interruption and have a chance to run full nayduck test

[bowenwang1996]

I investigated the CI failure and realized that it is not very easy to fix:

• We need to upgrade ethabi to 0.13 because of the issue with syn (update uint and ethereum-types rust-ethereum/ethabi#201)
• However, ethabi 0.13 depends on ethereum-types 0.10.0 and we use ethereum-types 0.6.2 in near-evm-runner because of the openethereum dependency.
• We cannot simply upgrade to the latest ethereum-types version because here

  nearcore/runtime/near-evm-runner/src/near_ext.rs

  Line 70 in e3cd9fc

  impl<'a> vm::Ext for NearExt<'a> {

  we implement an openethereum trait and therefore the types must match what they use.

I am not sure how to best proceed at this point. I feel like we can either pin some syn dependency or try having two different ethereum-types version in the same crate. @frol @artob any suggestions?

[bowenwang1996]

@frol also looks like there is some change to maybe actix runtime that causes trouble when we start a node:

thread 'main' panicked at '`spawn_local` called from outside of a `task::LocalSet`', /Users/bowenwang/.cargo/registry/src/github.com-1ecc6299db9ec823/tokio-1.1.0/src/ta\
sk/local.rs:303:18
stack backtrace:
   0: rust_begin_unwind
             at /rustc/91a79fb29ac78d057d04dbe86be13d5dcc64309a/library/std/src/panicking.rs:483
   1: core::panicking::panic_fmt
             at /rustc/91a79fb29ac78d057d04dbe86be13d5dcc64309a/library/core/src/panicking.rs:85
   2: core::option::expect_failed
             at /rustc/91a79fb29ac78d057d04dbe86be13d5dcc64309a/library/core/src/option.rs:1226
   3: core::option::Option<T>::expect
             at /Users/bowenwang/.rustup/toolchains/nightly-2020-10-08-x86_64-apple-darwin/lib/rustlib/src/rust/library/core/src/option.rs:346
   4: tokio::task::local::spawn_local::{{closure}}
             at /Users/bowenwang/.cargo/registry/src/github.com-1ecc6299db9ec823/tokio-1.1.0/src/task/local.rs:302
   5: tokio::macros::scoped_tls::ScopedKey<T>::with
             at /Users/bowenwang/.cargo/registry/src/github.com-1ecc6299db9ec823/tokio-1.1.0/src/macros/scoped_tls.rs:72
   6: tokio::task::local::spawn_local
             at /Users/bowenwang/.cargo/registry/src/github.com-1ecc6299db9ec823/tokio-1.1.0/src/task/local.rs:301
   7: actix_rt::arbiter::Arbiter::spawn
             at /Users/bowenwang/.cargo/registry/src/github.com-1ecc6299db9ec823/actix-rt-2.0.0-beta.2/src/arbiter.rs:162
   8: actix_rt::spawn
             at /Users/bowenwang/.cargo/registry/src/github.com-1ecc6299db9ec823/actix-rt-2.0.0-beta.2/src/lib.rs:35
   9: actix_http::client::pool::ConnectionPool<T,Io>::new
             at /Users/bowenwang/.cargo/registry/src/github.com-1ecc6299db9ec823/actix-http-3.0.0-beta.1/src/client/pool.rs:67
  10: actix_http::client::connector::Connector<T,U>::finish
             at /Users/bowenwang/.cargo/registry/src/github.com-1ecc6299db9ec823/actix-http-3.0.0-beta.1/src/client/connector.rs:344
  11: near_telemetry::TelemetryActor::new
             at /Users/bowenwang/NEAR/nearcore/chain/telemetry/src/lib.rs:50
  12: neard::start_with_config
             at /Users/bowenwang/NEAR/nearcore/neard/src/lib.rs:227
  13: neard::main
             at /Users/bowenwang/NEAR/nearcore/neard/src/main.rs:236
  14: core::ops::function::FnOnce::call_once
             at /Users/bowenwang/.rustup/toolchains/nightly-2020-10-08-x86_64-apple-darwin/lib/rustlib/src/rust/library/core/src/ops/function.rs:227
note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.
Panic in Arbiter thread.

[birchmd]

I made some changes to main.rs which seem to fix the panicked at spawn_local issue. I'm not 100% sure yet if it is correct, but the node seems to run ok with the change.

diff --git a/neard/src/lib.rs b/neard/src/lib.rs
index 92c3db80..18deb412 100644
--- a/neard/src/lib.rs
+++ b/neard/src/lib.rs
@@ -214,7 +214,6 @@ pub fn start_with_config(
     config: NearConfig,
 ) -> (Addr<ClientActor>, Addr<ViewClientActor>, Vec<Arbiter>) {
     let store = init_and_migrate_store(home_dir, &config);
-    near_actix_utils::init_stop_on_panic();

     let runtime = Arc::new(NightshadeRuntime::new(
         home_dir,
diff --git a/neard/src/main.rs b/neard/src/main.rs
index 814dfeb5..41ae7ce9 100644
--- a/neard/src/main.rs
+++ b/neard/src/main.rs
@@ -232,10 +232,11 @@ fn main() {
                 near_config.client_config.archive = true;
             }

-            let system = System::new("NEAR");
-            let (_, _, arbiters) = start_with_config(home_dir, near_config);
-            system.run().unwrap();
-            arbiters.into_iter().for_each(|mut a| a.join().unwrap());
+            System::builder().name("NEAR").stop_on_panic(true).run(move || {
+                start_with_config(home_dir, near_config);
+            }).unwrap();
         }
         ("unsafe_reset_data", Some(_args)) => {
             let store_path = get_store_path(home_dir);

[frol]

@bowenwang1996 I forked openethereum and upgraded the dependencies. I had also forked actix-extras and paperclip (we use it for RosettaRPC). Now, we have a single version of actix (0.11.0-beta.1), actix-web (4.0.0-beta.1), and tokio (1.1.1)!

@birchmd Thanks for the patch! I haven't yet confirmed if that fixes the issue, but it looks reasonable to me.

[frol]

New nayduck link: http://nayduck.eastus.cloudapp.azure.com:3000/#/run/1098