Re:Understanding US Power Outages – By Brian Potter
by rbanffy on Tuesday April 15, 2025 @06:17AM
(#43690953)
> I had a server with over 3 years of uptime
Dear God! Update that kernel!
Re:Understanding US Power Outages – By Brian Potter
by sgt on Tuesday April 15, 2025 @06:36AM
(#43691039)
3 years is nothing! I have a production server with 2400 days uptime. At this point getting to 3000 is more important than security.
Re:Understanding US Power Outages – By Brian Potter
by rbanffy on Tuesday April 15, 2025 @06:59AM
(#43691167)
What is it? An IBM z14? Counting PR/SM uptime is cheating.
Re:Understanding US Power Outages – By Brian Potter
by sgt on Tuesday April 15, 2025 @07:09AM
(#43691229)
Ubuntu 16.04 or something.
I've been getting a good taste of this one. If you want an excellent ~real time visualization of just how fragile these areas are, pull up the Entergy outage tracker the next time a hurricane rolls through Texas/Louisiana.
In my current neighborhood, approximately 90% of the homes have LNG standby generators. A typical outage in the area is at least 12 hours to resolve. With a catastrophic windstorm it takes upward of 2 weeks.
On the other end of the spectrum, I used to live close to the original Samsung factory in Austin. I was in an apartment that was fed with underground utility. I had a server with over 3 years of uptime and it was plugged directly into the wall. This was 2011-2014.
I think a lot of it boils down to your local infrastructure. If you pay really close attention to how the utilities are implemented at real estate transaction time, you can influence how much impact you would see during a disaster.
If you're in ERCOT and you absolutely cannot handle the idea of losing power ever again, buying something around Taylor, TX would probably be a good way to bunker up. You need to understand that the Koreans would never build another semiconductor factory in Texas if they didn't have incredible guarantees from the utility providers moving forward from that winter event.