If something goes terribly wrong and we can't start the node app after an update, we have no way to run any commands at the moment (changes to the entry point script are applied from within the node app, so those won't have any effect). The idea is to update the entry point script so that it executes patch.sh from the update archive, if it exists, which gives us a chance to apply fixes before node starts.
If something goes terribly wrong and we can't start the node app after an update, we have no way to run any commands at the moment (changes to the entry point script are applied from within the node app, so those won't have any effect). The idea is to update the entry point script so that it executes
patch.sh
from the update archive, if it exists, which gives us a chance to apply fixes before node starts.