I'm sorry... I tripped.
HN is up again
181–190 of 390 posts
Re: HN is up again
#182I’d love to learn more about HN hardware & software stack. I know old posts indicate it’s running on a low core count but high frequency Intel CPUs on FreeBSD and no database (just flat files). I wonder if it’s still the same.
Ah, that's easy.
HN is running on an old laptop from Viaweb.
Arc is running under the pg user and it's used as the process supervisor.
The actual web server is a VB app running on Linux through Wine.
The flat files have been migrated to an MS Access DB, also running through Wine.
Re: HN is up again
#183I've been noticing internet problems here and there all week, and was getting a little sketched out (is the heat waves, Russia, or alien attack:) so this really got me worried.
Re: HN is up again
#184Earlier quoted context omitted.
Yep. Same. It says a lot about the quality of HN, I think. Also, I can't remember the last time it was down. For a while, I thought there must be something wrong with my internet connection or DNS config or something.
Indeed. I find that it's so reliable and fast that I use it to check my internet connection. If I can't hit HN, then something's wrong on my end.
I assumed at first there was a problem with my mobile data connection, before I realized it was actually HN :)
Re: HN is up again
#185I don't think I had ever fully internalized how often I open this site throughout the day. Finish a task? HN. Got frustrated/stuck on a problem? HN break. Waiting for something to install/upload/compile/etc? HN. Needless to say I opened a new tab, typed "n", and hit enter countless times today before my brain caught up with my muscle memory.
This happened last couple of times I switched laptops - my old habit to visit "guardian.co.uk" by typing "guar" and hitting enter no longer works because I've now accidentally searched too many times for "guar" :D
Re: HN is up again
#186This was a rough one. I actually opened slashdot at one point.
Same
Early on /. was amazing! Remember Cmdrtaco it all out, often taking us for the ride?
Good times, frequently good discussion.
HN has been better for years now, was better at inception, for the most part.
/. has improved a bit. Good to see, or I caught it on a good day.
Re: HN is up again
#187HN is my default 'is the internet working' site and that genuinely threw me for a loop across multiple devices today dealing with hotspots while our power was out.
Re: HN is up again
#188Whenever I get frustrated by cloud complexity I wonder if its all worth it, as HN, stackoverflow, camelx3 etc are still on real servers. Maybe it is worth it after all.
I've wondered about this a lot recently as I've spent a lot of time recently fighting with Terraform, Atlantis, CircleCI, github and AWS. The first half of my career was deploying code to various UNIX machines. When there was a problem I could login to that machine and use various (common) tools to diagnose the issue. I may not have had root but I was at least able to gain insight into what was likely the culprit. Th…
Lol, same boat. Not sure if I'm the old wise guy who really knows his sh*t about what's important, or the old useless guy rambling and moaning in the corner.
Re: HN is up again
#189granted, this is probably the most exciting post here all day!
Re: HN is up again
#190Earlier quoted context omitted.
Yep: if you buy a pair disks together, there's a fair chance they'll both be from the same manufacturing batch, which correlates with disk defects.
This makes total sense but I've never heard of it. Is there any literature or writing about this phenomenon? I guess proper redundancy is having different brands of equipment also in some cases.
I also don't know about literature on this phenomenon, but i recall HP had two different SSD recalls because when the uptime counter rolled over, they would fail. That's not even load dependent, just did you get a batch and power them on all at the same time. Uptime is too high causing issues isn't that unusual for storage, unfortunately.
It's not always easy, but if you can, you want manufacturer diversity, batch diversity, maybe firmware version diversity[1], and power on time diversity. That adds a lot of variables if you need to track down issues though.
[1] you don't want to have versions with known issues that affect you, but it's helpful to have different versions to diagnose unknown issues.