Convenience and cost trumped everything, because no one understood (or believed there would be) consequences to handing over their data and habits to "free" services. Drug pusher model got them hooked, and now here we all are as a society...
-
Here's a blog post. There's a full research paper I don't have a link to on my phone I can find later, from the mod 00's. https://www.microsoft.com/en-us/research/blog/better-way-store-data/ …
I'll check out the paper once you get a chance to find it. I think you may be overestimating the prevalence of LRC, though. RS is arguably still the baseline standard and other approaches to solve the problem are around (e.g. SHEC in Ceph and Google's nested coding).
-
-
I'll admit I haven't looked at this much the last few years, so I'd believe that. I was under the impressions that there are derivative alternatives that solve the same problem: Rebuilding data with as little network activity as possible, because the disk sets are so distributed.
-
Here's the best one I can find right now. Look at Cheng Huang's entire research history tho. It's basically all dedicated to this problem.https://www.microsoft.com/en-us/research/publication/rethinking-erasure-codes-cloud-file-systems-minimizing-io-recovery-degraded-reads/ …
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.