er0k

The Problem

There is a longstanding systemd/networkd bug that causes our pet instances (AWS EC2 Ubuntu) to fall off the network, detailed in this gist.

To confirm if you are affected, look for messages like Could not set DHCPv4 address: Connection timed out in syslog by running grep 'Connection timed out' /var/log/syslog or zgrep 'Connection timed out' /var/log/syslog.*.gz on your pet.

If your pet is currently inaccessible, you'll need to restart it via the awscli by running:

aws --profile "$PROFILE" ec2 reboot-instances --instance-ids "$INSTANCE"

er0k

#!/bin/bash

# aws profile name for your pet
PROFILE="pet-role-yourname"
# aws instance ID
INSTANCE="i-00000abcd0000abcd"

case "$1" in
    start|stop|reboot)
        aws --profile "$PROFILE" ec2 "${1}"-instances --instance-ids "$INSTANCE"

er0k

#!/bin/bash

# https://wiki.debian.org/iPhone

# installs deps
# apt install libimobiledevice6 libimobiledevice-utils ifuse

# pair the phone. will prompt for passcode
idevicepair pair || exit 1

er0k

#!/usr/bin/env python

import time


def isAllowed(clientId, map) -> bool:
    now = round(time.time())

    if now in map:
        map[now].append(clientId)

er0k

curlie \
  -H "Authorization: Bearer $LINODE_TOKEN_RW" \
  -X POST -d "$(jo \
    image="linode/debian11" \
    root_pass="$(pwgen 128 1)" \
    authorized_keys="$(curl -s -H "Authorization: Bearer $LINODE_TOKEN_RW" 'https://api.linode.com/v4/profile/sshkeys' \
      | jq -r '.data[].ssh_key' \
      | jo -a)" \
    booted=true \
    label=zzz \