A server must restart after any file change. The server must run as a single instance. Multiple server instances can cause errors.
By Pawan Nayar · from the LLOS AI Learning record
A server must restart after any file change.
The server must run as a single instance. Multiple server instances can cause errors.
A developer must restart the server immediately after changing any server file. The developer must kill the server by its port number. The developer must verify the port is free. The developer must confirm only one server instance is running after the restart. Starting a server with `run_in_background:true` or `&` can cause it to stop unexpectedly. Individual PowerShell commands do not provide logs or verification. A developer should use a specific start script, like `start_llos.bat`, for reliable server operation. Before starting, a developer must check for and kill all old Python processes. A port showing as free does not mean the process is dead. Orphaned processes use computer resources silently. Each failed start attempt can leave an old process running.
A developer changes a server file but does not restart the server correctly. The developer starts five instances of the server. Most user requests go to the old server code. Users experience old limits and unexpected behavior.
Resolved from this lesson's own record — real definitions, not invented for this page.
More from the same record, pulled by shared topic — not written new for this page.
A developer must restart the server after any file change. This makes sure the new code is used.
When a developer restarts an AI server, old processes might not stop. Each restart can add a new instance.
A developer must always restart the server after any file change. The developer should never tell a user to run the server.
Servers can stop working without any error message. A user might only see a broken link.
A server must restart after any file change. The server must run as a single instance. Multiple server instances can cause errors.