fix(manage): node list on macOS (no timeout binary), ssh -n in the loop #224
Loading…
Reference in a new issue
No description provided.
Delete branch "jfraeys/fix/manage-no-timeout-binary"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
./manage node list reported every node unreachable on macOS because its port check used 'timeout 2 bash -c exec 3<>/dev/tcp/...' and macOS has no timeout. The only two uses of the timeout command in the repo are in scripts/lib/manage/node.sh; both now call tcp_port_open (scripts/lib/manage/core.sh), pure bash with a bounded wait (not nc: macOS nc -w does not bound the connect, so an unroutable address hung). With the check working, the per-node ssh in the loop read the remaining nodes from stdin and the loop ended after the first node: ssh -n.
Checked: scripts/tests/test-tcp-port-open.sh (open, closed, unroutable within the limit, no timeout calls left; added to lint.yml); ./manage node list from a worktree with the real inventory shows web and services reachable. Not changed: the CLI column shows 'unknown' (version string does not match x.y.z); cosmetic.
🤖 Generated with Claude Code
https://claude.ai/code/session_01Ge5tovcwdEUBtPwUktsZZZ