Serious question: how much of this could have either had been prevented, or restored much more quickly if they were using ZFS with proper parity, checksuming, snapshotting, and sending (backups)? This really is the one-size-fits-all storage solution at this point.
Here, I found an entire web site dedicated to helping find these correlations! http://www.tylervigen.com/spur...
Maybe nobody emails them specifically? I still get ~7,000 junk emails per month (caught by spam filters), compared to maybe 200-500 legit messages.
I wanted to make a 2G (bandwidth) vs 4G (cell network) joke, but decided to pop up with some actual useful information instead.
Currently Comcast has this in their contract:
"For upgrading from Performance to Blast!®, ranges based on area, from $10.00 to $12.00 more per month (subject to change)"
This is an additional service charge ON TOP OF the higher price already charged to get faster internet service. That's right, they're advertising one price, and then tacking on additional fees for that bandwidth. Just for having faster bandwidth, you're forced to pay an additional $10+ on top of the already higher rate.
So, what is the premium "Blast" or "2G" fees going to be with this new service? Sure, the base rate is already an outrageous $300... but this is Comcast, who is actively tacking on hidden fees to customers outside of the usual regulatory fees.
I can't even brain after read that summary...
Unlicense, and done. http://unlicense.org/
When buying a company, part of what you're buying are their assets, both tangible and intangible. This is NOT exclusive to just modern internet companies. Go anywhere as far back in history as you'd like. When one company buys another, why would they NOT transfer over customer account records?
Just imagine the inverse for a second... The company you're doing business with gets bought out, but are not allowed to transfer over their records. You walk into that business the next day, and before you can even do anything, you are greeted with "SORRY, new ownership, you have to start over your account with us from scratch"
Now take that example, and apply it to:
ISP merger. "SORRY you have to re-sign up for your internet access"
Bank: "SORRY, you need to sign up for a new checking/savings/credit account"
Clubs: "SORRY, you're not a member anymore!"
Mortgage: "SORRY, you don't own your house anymore"
Looks like Chromecast has gone the way of Google Chrome: Arbitrary and random A/B testing that you're never notified of, and no way to opt out of.
This seriously pisses me the fuck off with Chome. The browser works great on like 20 machines, and then fucks up on one. You think it is the machine's fault, until you dig and dig and dig into vague forum posts on Google's boards. Then it turns out to be a hidden A/B test, where you have to go into the hidden Chrome settings to force enable/disable some very specific feature to get out of the that one and only that one particular test.
This is EXACTLY what happened to my primary development machine. Chrome had a hidden A/B test for ASync DNS requests. This feature is bugged to shit and back during the test. It would lock the entire browser session (all tabs) for 30-60 seconds at a time while making only certain DNS requests.
Another example is with the internal cache system. There was a bug for a while which would also lock up Chrome for 30-60 seconds at a time just waiting to see if a URL resource is locally cached. There was no fix for this that I could find. My resolution was eventually to have the installers handy for both Release and Beta Chrome laying around. Sometimes Release was the broken build, sometimes Beta was the broken build. So when shit got fucked up, I'd just toggle between the builds.
Yeah, the science hasn't even been tested yet... So, what is this, just hopeful, wishful thinking? What is the false positive rate? More importantly, what other chemicals trigger a false positive?
Remember that "date rape testing fingernail polish" that went super viral? Awesome in theory, horrible in practice. Milk causes a false positive with it. How many drinks nowadays contain some form of milk? Rumchata and White Russians both come to mind instantly. [in before "in soviet russia" joke]
This isn't possible on all systems. Thanks to bullshit in the tech industry, a bunch of us are stuck on laptops with a "720p" display, because for years the industry thought it would be awesome to lock everyone down to this size, despite having 1600x1200 become semi common places years prior.
TLDR: It shouldn't be up to the use to fix bad design by designers. And these titles getting cut off is getting seriously fucking annoying.
Samsung is not the one at fault with the drivers here. The example stated as USB 3.0 ports. How many computer OEMs make their own USB chipsets? My guess would be none. The source the chips from other vendors, and then those chips simply register as another PCI device attached to the system bus. This is also extremely true for NICs, how many onboard NICs are Realtek, not Dell/Asus/Acer/Samsung/whatever? WU treats these devices as individual devices, not part of a total package computer from an OEM.
And why does this "work" for other OEMs? Read other user comments here. Plenty people complaining about fixing WU driver update issues. It IS a problem.
For example: AMD pushed an update out that broke the SATA drivers for their motherboard. This was exceptionally annoying in that I use a dedicated storage controller which also acts as my boot device, so I wasn't even using my SATA ports. The bug was so bad it still prevented Windows from booting at all. (this was maybe a year or two ago now)
Good thing that was already solved back in 2001. https://tools.ietf.org/html/rf...
Having a quality backup solution isn't all that hard these days.
On-site file server with a ZFS RAID-Z (2/3) storage pool. Frequent snapshots of data (hourly?). Occasional ZFS Sends to offsite location over VPN (nightly?)
Occasional ZFS scrubbing, which validates block level data against hashes rather than just a basic checksum/parity bit/SMART check. (weekly/monthly?)
Single drive failure? Just replace it, nothing is down.
Multi-drive failure? Depends on your RAID-Z level, but possibly still nothing down, and just replace the failed drives.
Accidentally modify/delete something? Just mount a snapshot and recover.
Entire storage server goes offline? Set a new one up with fresh storage, and just ZFS Send to it from the off-site server.
All of this is possibly from something as simplified as FreeNAS, or can be baked into a more robust solution as well.
"MongoDB is webscale"
For those that have not seen it yet: https://www.youtube.com/watch?...
UNIX is hot. It's more than hot. It's steaming. It's quicksilver lightning with a laserbeam kicker. -- Michael Jay Tucker
