From c8a839aeffa76b0b3325d642ee38bba2595170b4 Mon Sep 17 00:00:00 2001
From: Wen <113942165+wen-coding@users.noreply.github.com>
Date: Sat, 24 Aug 2024 19:32:11 -0700
Subject: [PATCH] Enable wen restart from command line and update comments.
 (#2640)

Enable wen_restart from command line and update comments.
---
 validator/src/cli.rs  | 19 ++++++++++---------
 validator/src/main.rs |  6 ++++++
 2 files changed, 16 insertions(+), 9 deletions(-)

diff --git a/validator/src/cli.rs b/validator/src/cli.rs
index 9f4276d8d67ee0..38410aaeabd7ab 100644
--- a/validator/src/cli.rs
+++ b/validator/src/cli.rs
@@ -1580,7 +1580,9 @@ pub fn app<'a>(version: &'a str, default_args: &'a DefaultArgs) -> App<'a, 'a> {
                 .required(false)
                 .conflicts_with("wait_for_supermajority")
                 .help(
-                    "When specified, the validator will enter Wen Restart mode which \
+                    "Only used during coordinated cluster restarts.\
+                    \n\n\
+                    When specified, the validator will enter Wen Restart mode which \
                     pauses normal activity. Validators in this mode will gossip their last \
                     vote to reach consensus on a safe restart slot and repair all blocks \
                     on the selected fork. The safe slot will be a descendant of the latest \
@@ -1588,16 +1590,15 @@ pub fn app<'a>(version: &'a str, default_args: &'a DefaultArgs) -> App<'a, 'a> {
                     optimistically confirmed slots. \
                     \n\n\
                     The progress in this mode will be saved in the file location provided. \
-                    If consensus is reached, the validator will automatically exit and then \
-                    execute wait_for_supermajority logic so the cluster will resume execution. \
-                    The progress file will be kept around for future debugging. \
-                    \n\n\
-                    After the cluster resumes normal operation, the validator arguments can \
-                    be adjusted to remove --wen_restart and update expected_shred_version to \
-                    the new shred_version agreed on in the consensus. \
+                    If consensus is reached, the validator will automatically exit with 200 \
+                    status code. Then the operators are expected to restart the validator \
+                    with --wait_for_supermajority and other arguments (including new shred_version, \
+                    supermajority slot, and bankhash) given in the error log before the exit so \
+                    the cluster will resume execution. The progress file will be kept around \
+                    for future debugging. \
                     \n\n\
                     If wen_restart fails, refer to the progress file (in proto3 format) for \
-                    further debugging.",
+                    further debugging and watch the discord channel for instructions.",
                 ),
         )
         .args(&thread_args(&default_args.thread_args))
diff --git a/validator/src/main.rs b/validator/src/main.rs
index 87b6a44cd9666e..c61b0f6d3ec87e 100644
--- a/validator/src/main.rs
+++ b/validator/src/main.rs
@@ -1523,6 +1523,7 @@ pub fn main() {
         replay_transactions_threads,
         delay_leader_block_for_pending_fork: matches
             .is_present("delay_leader_block_for_pending_fork"),
+        wen_restart_proto_path: value_t!(matches, "wen_restart", PathBuf).ok(),
         ..ValidatorConfig::default()
     };
 
@@ -1965,6 +1966,11 @@ pub fn main() {
     let mut node = Node::new_with_external_ip(&identity_keypair.pubkey(), node_config);
 
     if restricted_repair_only_mode {
+        if validator_config.wen_restart_proto_path.is_some() {
+            error!("--restricted-repair-only-mode is not compatible with --wen_restart");
+            exit(1);
+        }
+
         // When in --restricted_repair_only_mode is enabled only the gossip and repair ports
         // need to be reachable by the entrypoint to respond to gossip pull requests and repair
         // requests initiated by the node.  All other ports are unused.