webdekeepalive.config.sample

# webdekeepalive.py configuration (global settings for all accounts)
# Copy to 'webdekeepalive.config' and adjust. Keep the file private: chmod 600 webdekeepalive.config
# The file is optional, without it the defaults shown here apply.
# All paths are relative to the project directory.

# ============================================================================
# Directories and files
# ============================================================================
[DIRECTORIES]
# Directory with the account files (<name>.env)
accounts_dir = secrets
# Directory for the log file and the lock file
logs_dir = logs

[FILES]
log_file = webdekeepalive.log
# Prevents two runs at the same time
lock_file = webdekeepalive.lock

# ============================================================================
# Timing
# ============================================================================
[TIMING]
# Pause in seconds between two accounts when one run handles several accounts
# (the systemd timers run every account on its own, see SCHEDULE in the account files)
delay_between_accounts = 10
# When another run holds the lock (e.g. two account timers catching up after a reboot),
# wait up to this many seconds for it to finish instead of skipping the run
lock_wait = 900

# ============================================================================
# IMAP login (the actual keep-alive)
# ============================================================================
[IMAP]
# Timeout in seconds for connecting and for each IMAP command.
# Can be overridden per account with ACCOUNT_IMAP_TIMEOUT.
timeout = 30
# Attempts for network problems (WLAN/DNS hiccups, timeouts, dropped connections) with
# exponential backoff: base_delay, 2x, 4x, ... capped at max_delay (30s, 60s, 120s).
# A rejected password is never retried, so the account is not locked.
attempts = 4
base_delay = 30
max_delay = 300

# ============================================================================
# SMTP (optional report mail, see MAIL_SEND in the account files)
# ============================================================================
[SMTP]
timeout = 30
# Only connecting and login are retried, so a report mail is never sent twice
attempts = 2
base_delay = 30
max_delay = 300

# ============================================================================
# Logging
# ============================================================================
[LOGGING]
# Write a log file. Set to false to log to stderr only (journald when run by the
# systemd timer), which avoids writes to the SD card completely.
file_enabled = true
# DEBUG, INFO, WARNING, ERROR. INFO writes about 5 lines per account and run.
level = INFO
# Rotation: max size per file in bytes and number of old files kept.
# 1 MB with 2 backups = at most 3 MB on disk.
max_bytes = 1048576
backup_count = 2

# ============================================================================
# Heartbeat (e.g. Uptime Kuma push monitor)
# ============================================================================
# After every run the script calls this URL with status=up (all accounts logged in)
# or status=down (at least one failed, configuration error, crash or SIGTERM) and a
# short message. If the Pi or the script stops running, the push is missing and the
# monitor goes down on its own. A heartbeat per account is possible with HEARTBEAT_URL
# in the account file. See docs/deploy_git.md, section "Uptime Kuma".
[HEARTBEAT]
enabled = false
# Push URL from Uptime Kuma. Parameters copied along (?status=up&msg=OK&ping=) are
# replaced by the script, so the URL can be pasted as shown in Uptime Kuma.
url = https://uptime.example.com/api/push/your_push_token
# Request timeout in seconds and retries for connection errors, 429 and 5xx
timeout = 10
retries = 3