Follow

Starting this instance on a server with "some space and an IPv4 address available" was a naive thing. I'll figure out how to move to the bigger machine next week. So far everything worked with zero downtime, but moving to a bigger server might mean ~15 minutes unavailable when I either have to move a subnet or change IP address – or both.

· · Web · 6 · 5 · 13

@admin Bitte dran denken: Das Wartungsfenster auf Kipp zu lassen, erhöht die Energiekosten! Schließen Sie das Wartungsfender nach Abschluss der Arbeiten.

@mattias That's a really bad german play with words on "maintenance window". Yes, I'll close it when done.

@admin reminds me of when I did sys admin. Crank down the DNS TTL, wait, then switch and crank the TTL back up again. Good luck and thank you.

@chrisgerhard Routing issues are far worse than TTL. Basically: No one honors DNS TTL below 600s, so you kno and you can tell customers that 10 minutes here and ten minutes there can add up…but propagating updated routing tables for the same is kinda Las Vegas. Despite both servers are in Falkenstein a few hundred metres apart.

@admin 10 minutes of port forwarding or 10 minute outage. As a user I'd go with the outage.

@admin thank you very much. Is there a way we can financially support the bigger machine (ideally via paypal)? Didn't found any note on that topic.

@admin thanks for all the efforts 👍 and thanks for letting me find a little home on your system

Sign in to participate in the conversation
toot.bike

A Mastodon instance aimed at (but not limited to) the cycling community.