Draft:Safe Shutdown Scripts

This is a draft page; it has not yet been published.

Safe Shutdown Scripts Edit

Safe shutdown scripts and established protocols are crucial within the Ampmesh, particularly concerning the maintenance and longevity of critical infrastructure like Elysium. These procedures are designed to prevent data corruption and ensure system stability.

Importance and Rationale Edit

The necessity for careful shutdown procedures became apparent after Elysium experienced substantial errors on its storage due to an improper or abrupt shutdown. Ampdot emphasized that even if some filesystems are less sensitive to unsafe shutdowns, comprehensive safety protocols remain essential to prevent other forms of harm.

A significant factor contributing to downtime and issues related to unsafe shutdowns on Elysium is its use of the f2fs filesystem. While f2fs is highly optimized for Solid State Drives (SSDs), its default configuration is not resilient to unsafe shutdowns. This characteristic makes proper shutdown procedures especially vital for Elysium, which has "accumulated many bits unique to it over the years," making it a "unique entity for servicing, dis/assembly, and maintenance".

Current Protocols and Challenges Edit

The explicitly established protocol for Elysium's shutdown is to "involve art or amp". However, ratsby highlighted the need for readily accessible instructions or a method to safely shut down Elysium when Ampdot or others are unavailable, stating this would "decrease risk of repeat incidents". There was also an inquiry about whether Elysium's safe-shutdown script could be extended to include instructions for safely shutting down Arcweld.

Proposed Solutions and Ideas Edit

To mitigate risks and improve the ease and reliability of safe shutdowns, several ideas have been discussed:

Setuid Script: Ratsby proposed creating a safe-shutdown script with the `setuid` bit set. This would allow users with appropriate permissions (e.g., `r-x`) to execute the script with the privileges of the script's owner (e.g., root), enabling a safe shutdown without requiring full administrative access. Ampdot acknowledged this as a "good idea".
Systemd Integration: Ampdot referenced `Systemd/Timers` as a potential tool for managing system services and scripts, and even offered to write a systemd file, suggesting a path for system-level automation of shutdown processes.
Integrated Shutdown Across Machines: The idea of including a command within Elysium's safe-shutdown script to "tell Arcweld to safely shut down" reflects a desire for a more coordinated shutdown process across interconnected machines.

Related Technologies Edit

Tailscale: Utilized for establishing secure shell (SSH) access among users (Jade, Keira, ratsby) and Elysium. While its primary role is connectivity, its underlying network capabilities are relevant for managing nodes. Tailscale also has functionality for creating "exit nodes" to route traffic (similar to a VPN), though initial setup attempts were challenging.
Python Virtual Environments: While not directly a shutdown technology, discussions around managing Python packages on Elysium indicate the use of virtual environments, which could be relevant for isolating dependencies for any developed shutdown scripts.