nonsense

#20973 local reproduction follow-up

Issue: reth-optimism-node::it p2p_version::peers_negotiate_eth_69 intermittently exits with SIGSEGV after the test body reports ok.

Date: 2026-05-22

Summary

The original findings are directionally correct: upstream reth changed the payload builder from a Tokio task to an untracked OS thread, and op-reth inherits that path. I could not reproduce the exact SIGSEGV locally, including on native x86_64 Ubuntu inside Docker, but I did reproduce the lifecycle hazard directly by patching reth to widen the race window. The patched run shows orphan payload-builder OS threads touching a Tokio runtime after shutdown has begun.

nonsense

SDM / Flashblocks Workspace Change Notes

Date: 2026-05-07

Scope

This note covers the current committed workspace in:

• Main repo: /Users/nonsenseop/code/src/github.com/ethereum-optimism/optimism4
• Branch: nonsense/sdm/flashblocks

nonsense

#!/usr/bin/zsh

sectors=`lotus-miner sectors list | grep Available | grep -v "24 weeks" | awk '{print $1}' | paste -s -d' ' | tr -d '"'`

sectors_array=(`echo ${sectors}`);

epoch=`lotus status | grep Sync | awk '{print $3}'`

nextepoch=$((1555200+$epoch))

nonsense

gist for testground taas alpha

nonsense

## Global Docker image parameters
## Please, note that this will override the image parameters, including dependencies, configured to use the global value
## Current available global Docker image parameters: imageRegistry, imagePullSecrets and storageClass
##
# global:
#   imageRegistry: myRegistryName
#   imagePullSecrets:
#     - myRegistryKeySecretName
#   storageClass: myStorageClass

nonsense

c5.2xlarge - 8 vCPU ; 16 GB RAM

pod_resource_cpu = "300m"

pod_resource_memory = "300Mi"

30 c5.2xlarge worker nodes

500 testplan instances

nonsense

pod_resource_cpu = "300m" pod_resource_memory = "300Mi"

Feb 26 14:06:25.843143  INFO    all testplan instances in `Running` state       {"runner": "cluster:k8s", "run_id": "10f3e9470e0b", "took": "29.898906571s"}
Feb 26 14:06:47.974674  INFO    all testplan instances networks initialised     {"runner": "cluster:k8s", "run_id": "10f3e9470e0b", "took": "52.030438523s"}

nonsense

package main

import (
	"flag"
	"fmt"
	"sync"
	"time"

	"github.com/go-redis/redis/v7"
)

nonsense

cleandatadir,0
swarm-private-0,bc63781830079269fcdc7ec3be8c9f0a6995d7cca7e463cb64e27a664f33dc69
swarm-private-1,7b963a30587ea895e8afe32e1498da3dd3ad21240c0909228788520589e6c0c5
swarm-private-2,9cecdf34de9ac2312916c3a40c3f27a4bf83f660a3d12dd8fc14733db1539db8
swarm-private-3,9e3bdf8c44c0ca7da40ff9e726a06059d486f0ecb6459e1efb4d395209a02e91
swarm-private-4,c1c9d91460caba2479c6029f71431a054064745b5188f391017a36a80e89d571
swarm-private-5,860683b702632a8c990641c89d82f119ed801f3111bd7d202aee9a4234d066e4
swarm-private-6,bfa4ec2d0133addf2efe141caf19498eef0f616127b605f711625c904e7435a7
swarm-private-7,b24b1f52272f0f94951aaa7c0942528d0d978efda7b56a8c583d5d5ccf01a3ec
swarm-private-8,d7ad93082cf470cefde1df5c51aecb179d7d92d87b287683d8f0bbca12b40760

nonsense

#!/bin/bash -xe
 
echo '{ "log-driver": "json-file", "log-opts": { "max-size": "20m", "max-file": "10" }}' > /etc/docker/daemon.json
systemctl restart docker
 
CA_CERTIFICATE_DIRECTORY=/etc/kubernetes/pki
CA_CERTIFICATE_FILE_PATH=$CA_CERTIFICATE_DIRECTORY/ca.crt
mkdir -p $CA_CERTIFICATE_DIRECTORY
echo "${aws_eks_cluster.cluster.certificate_authority.0.data}" | base64 -d >  $CA_CERTIFICATE_FILE_PATH
INTERNAL_IP=$(curl -s http://169.254.169.254/latest/meta-data/local-ipv4)