3 ms·
Unlikely; a clock running at 30 Hz overflows a 32 bit (unsigned) integer in 39,768 hours, while the HN disk failed after at least 39,984 hours [1], and the vend
by Denvercoder9 4y ago
Unlikely; a clock running at 30 Hz overflows a 32 bit (unsigned) integer in 39,768 hours, while the HN disk failed after at least 39,984 hours [1], and the vendors wouldn't issues warnings about 40,000 hours if it actually fails about 230 hours before that.
[1] https://news.ycombinator.com/item?id=32031428 https://news.ycombinator.com/item?id=32031428
- abrax3141 4y agoNot sure about the mfgr- could be rounding for memory saliency. . . but re HN they’re counting clock time not power on time. Could have easily been turned off total 10 days for maintenance etc. (and good routine could have both servers off about the same no hrs.) That’s only 0.6% downtime.
- Denvercoder9 4y ago> Not sure about the mfgr- could be rounding for memory saliency. . . In that case, I would expect them to be rounding down, not up, and to at least mention the exact number somewhere. > but re HN they’re counting clock time not power on time. Could have easily been turned off total 10 days for maintenance etc. I seriously doubt HN has been down for 10 days over the last five years. As discussed in the other thread I linked, this has been one of the longest HN outages.
- abrax3141 4y agoYeah, no, I don’t mean crashed altogether, just off for routine maintenance on a rotating basis. But anyway, no sense in prolonging this. Anyway, agree that 1% seems high.