website maintainance update

some (although literally speaking none since nobody cares, but im posting this anyway) may have noticed my website was going up and down repeatedly over the past 24 hours. there were two internal errors the couldve become a huge issue if i didnt catch them early (which i did because my internet went down and that exposed them).

one; apparently you have to manually tell debian to reconnect to the internet if the connection is interrupted, even if its connected through internet. in practice, this means that if the internet went out even for a split second, the server would instantly give up connecting to the internet at all until its restarted, which would obviously be a huge issue because then i couldnt even remotely restart it if i was away from home. i was able to fix this by... well, just telling nmcli to autoconnect.

two; the search engine part of the website, which is hosted through a separate network management tool from the website (docker) wasn't automatically restarted when the server reboots or when its interrupted for any reason. as a result, you would only fetch an internal error if you tried to connect to the search engine after a restart, and the only way i was able to fix this was manually removing and restarting the container running the search engine. i solved the issue by using a flag that exists that tells the search engine's container to automatically restart it any time it stops, unless its manually stopped via a direct command.

website should be working again. i will be posting maintainance notices for the website here for future reference.


0 Kudos

Comments

Displaying 0 of 0 comments ( View all | Add Comment )