+- HP Forums (https://www.hpmuseum.org/forum)
+-- Forum: Not HP Calculators (/forum-7.html)
+--- Forum: Forum Issues and Administration (/forum-19.html)
+--- Thread: How long was the forum down today (8/12/2019) (/thread-13444.html)
How long was the forum down today (8/12/2019) - Gene - 08-13-2019 12:23 AM
I found that it was down for at least 2 hours from 5pm to 7pm my local time today. Back up of course.
Anyone else know it was down today?
RE: How long was the forum down today (8/12/2019) - Dave Hicks - 08-13-2019 12:34 AM
I got an email about it when I got home today. The email was sent a few hours ago.
I had to use phpMyAdmin to repair the sessions table. Maybe some myBB bug or hacking or maybe this??
https://www.wnycstudios.org/story/bit-flip
RE: How long was the forum down today (8/12/2019) - DA74254 - 08-13-2019 04:22 AM
I noticed it was down with SQL errors.
Only tried a couple of times within a few minutes around 23:00 EEDT
RE: How long was the forum down today (8/12/2019) - jlind - 08-13-2019 06:20 AM
(08-13-2019 12:34 AM)Dave Hicks Wrote: I got an email about it when I got home today. The email was sent a few hours ago.
I had to use phpMyAdmin to repair the sessions table. Maybe some myBB bug or hacking or maybe this??
https://www.wnycstudios.org/story/bit-flip
Dave,
Thanks for bringing it back up.
Regarding the "bit flip", that's always a possibility. Chaos Theory at work with Murphy's Law. It's the reason satellites have additional bits for not only error-detection, but error-correction as they routinely have bits flipped from the significantly higher levels of radiation they're subjected to. Been decades, but IIRC, error correction requires two additional bits per byte, not one. They could be hardened for radiation, but the weight of the shielding required would make their launch infeasible and it wouldn't eliminate flipped bits, merely reduce the probability. It's easier to do error correction. I'm dismayed they didn't have error detection and correction embedded in their voting system given how critical it was and the consequences of election results that lacked veracity. Someone blew it with the Design FMEA and Risk Analysis.
Thanks,
John