add support for core dumps

[chetanyb]

Summary

Adds opt-in core dump support for debugging crashed nodes.

• New --coreDumps flag for local and Ansible deployments
• Granular targeting: "all", specific nodes (zeam_0,ream_0), or client types (zeam,qlean)
• Core dumps persist in <data_dir>/<node_name>/

Changes

Local deployment (parse-env.sh, spin-node.sh):

• Parse --coreDumps flag
• Add should_enable_core_dumps() helper for matching logic
• Apply --init --ulimit core=-1 --workdir /data to matching containers

Ansible deployment (run-ansible.sh, 7 role files):

• Pass --coreDumps through as enable_core_dumps extra var
• Add Jinja2 conditional to all client roles (zeam, ream, qlean, lighthouse, lantern, grandine, ethlambda)

Usage

Local

NETWORK_DIR=local-devnet ./spin-node.sh --node all --coreDumps all

Ansible

NETWORK_DIR=local-devnet ./spin-node.sh --node all --deploymentMode ansible --coreDumps zeam,ream

[ch4r10t33r]

Are we only looking to support coredump for the docker images only (and not binaries)? If so, can we update the README.md to reflect this?

[ch4r10t33r]

Looks good.