{"version":"2.1.246","anchor":"background-daemon-start-failures-are-classified-and-a-sleepi","canonical_anchor":"background-daemon-start-failures-are-classified-and-a-sleepi","heading":"Background daemon start failures are classified, and a sleeping machine gets one retry","tier":"notice","area":"Background Daemon","scope":null,"heads_up":null,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.246\/e\/background-daemon-start-failures-are-classified-and-a-sleepi","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.246","markdown":"### Background daemon start failures are classified, and a sleeping machine gets one retry\n\nBackground daemon start failures now name a cause, and a sleeping machine gets one retry.\n\n**What**\n\nWhen the background daemon fails to start, the failure now carries a cause: wrapper, ask, spawn, timeout, exited, or the operating system's spawn error code. A child that died is reported separately from one that simply never answered. If the wall clock jumped during a cold start, typically because the machine slept, and the daemon is still unreachable, the whole start is retried once.\n\n**Details**\n\n- Telemetry gains `child_exited`, `exit_code`, `exit_signal`, `budget_extended` and `clock_jump_retry`.\n\n**Evidence**\n\n`retrying the whole start once`\n\n- Area: Background Daemon\n- Tier: You'll notice\n- Useful: 3\/5\n- Signal: 2\/5"}