That's simply untrue. Your annoyance with what is vs isn't does not affect truth.I wish nerds would stop trying to gate keep what IS and what ISN'T. So damn condescending. Not everyone can have a million backup copies of their files. ANY copy of your most important files that can be easily restored is a backup.
Well, they tried to introduce the AI driven Recall recently that'd back up literally abso-****ing-lutely everything one ever did, ever, probably with a cloud data store involved too!Tangentially: What do smart lazy people do for Windoze backups? Is there anything as comprehensive and effortless as Time Machine is for macOS?
That's fair enough. I recognize I'm dealing with a relatively small sample size. It's important to note, though, unless you're extremely careful about contact details, they probably know who you are, too, which may skew your experience somewhat. It's a little like how I share a name with a major political donor ion my state and when I call my elected officials, I sometimes get put right through because staff assumes I'm a major donor. That shouldn't happen but it does.IME, you're not wrong—I can't say I've ever had a 100% positive experience with Synology support, and it's generally because they are super cagey in what information they're willing to disclose to the user about what's going on under the hood. Even the support dumps you send them are encrypted.
Hasn't stopped me from owning a string of syno boxes for years, including my current one, because i just can't be arsed to build a standalone NAS myself, but I do wish their support sucked less.
That's the first I've encountered such a claim. Do you have a source for this?They are, actually. HDD heads deteriorate when writing.
SHR works well in my experience on 4 drives. They use BTRFS on top of it (only the good bits, not the bits of btrfs that are suicidal to use). I've not had drive failures, but I have tested it by taking drives out (cue a lot of beeping and complaining but no problem rebuilding the array afterwards). It also scales up nicely; you just sequentially put bigger disks in, and the pool / volume simply grows.Funnily enough, Synology offers their own solution for that on their larger NAS - SHR. https://kb.synology.com/en-us/DSM/tutorial/What_is_Synology_Hybrid_RAID_SHR
I've used it (though on just a 2 disk array, so not really testing a lot), and was able to restore when one drive failed after a lightning strike. YMMV, not a backup, just a variation on RAID (as Synology itself says at the end of that article), etc.
I guess I should clarify - energy assist HDD heads deteriorate from writing. Older heads don't, but they can mechanically wear out from constant seeking.That's the first I've encountered such a claim. Do you have a source for this?
Because something doesn't cover the "smoking hole" scenario doesn't make it not a backup. However, one SHOULD plan for the smoking hole, I agree, if you really don't want to lose data.That's simply untrue. Your annoyance with what is vs isn't does not affect truth.
You can say, for example, that a nearline-style checkpoint/snapshot of your data is a "backup" and that'd be colloquially forgivable, but "colloquially forgivable" is not "technically correct."
Your on-box snapshots won't help you if your home (where the NAS we're talking about is ostensibly located) burns down, or if someone breaks in and steals your NAS. Backups must by definition be recoverable in what's commonly called the "smoking hole" scenario, where the primary location is reduced to a smoking hole in the ground by a fire or an asteroid or whatever.
This is why you'll hear "nerds" say things like "RAID is not a backup," and "checkpoints are not a backup." They're not. You can rage that words don't mean what you want them to mean—and I'd normally be right there with you, because linguistic prescriptivism is evil and stupid—but in this case, it turns out that common usage really does reflect the correct definition.
If there's solace to take here, it's that a lot of very smart people have thought about these things for a long time, and there really are best practices to adhere to when backing up data. This is a good thing. You're free to do you when it comes to your own data, but don't try to play the "hurp durp words can mean what I want" card in this particular case. You're dead-ass wrong, and there's a mountain of learned experience out there for you to examine if you want detailed explanations why.
(The problem might be that I'm one of the "nerds" you're calling out, but as a former enterprise architect responsible for the lifecycle of hundreds of PB, I am confident that I know more about this issue than you do. Bring the slings and arrows, I can take them.)
still need a source - never heard of heads (old or new) wearing out.I guess I should clarify - energy assist HDD heads deteriorate from writing. Older heads don't, but they can mechanically wear out from constant seeking.
So ... trust them to keep the service active instead of Google or Dropbox? You're just shifting dependencies. I do hope it still runs standalone open protocols.
3-2-1 absolutely does cover the claimed scenario. I was responding specifically to quoted user's implication that Kevin setting the terms of the conversation by laying out 3-2-1 was just "nerds" "trying to gate keep what IS and what ISN'T."Because something doesn't cover the "smoking hole" scenario doesn't make it not a backup. However, one SHOULD plan for the smoking hole, I agree, if you really don't want to lose data.
But I don't understand why 3-2-1 isn't a backup, as is claimed? 3-2-1 covers the smoking hole because one of the copies is offsite, on a server, one copy is onsite, on a secondary device, and one copy is on your local hard drive (or maybe one copy is on a second external device like a USB stick, BD-ROM or whatever.
How is that not good coverage for the smoking hole?
That's still not a source. And while heads wear out over a period of use, that's very different from the fact that SSDs writing data is actively destructive to the memory storage. It is not a matter of the device will eventually succumb to entropy. It is literally being slightly destroyed in the act of being written and this is inherent to the design of the storage medium. Sure, you get quite a few writes with decent brands and I'm not saying otherwise. It's simply that any storage of that sort has an inherently limited lifespan that is separate from the entropic reality that nothing lasts forever.I guess I should clarify - energy assist HDD heads deteriorate from writing. Older heads don't, but they can mechanically wear out from constant seeking.
Thank you for this Blade Runner moment ... truly, this backup discussion has reached a cinematic levelI've seen things you people wouldn't believe...
No, this is an industry mentality that sees tens of thousands of devices which is where you see the issues creep up.
You can believe me or not, but it doesn't change reality.So until you cite an actual source for write heads being damaged while writing data to spinning HDDs, I'm going to call your claim inaccurate
Better than nothing, at least. But I feel like you're bundling a bunch of conjoined points of failure into one box. One hard drive fails? Data's lost. Sounds like you have different drives for different purposes so that alone any one drive failing isn't catastrophic, just a slice of your storage/backup/version history, but it is irrevocable.I'm looking at this JBOD device to make storage and access easier now that I've moved to a laptop. https://www.amazon.com/gp/aw/d/B07Y4F5SCK .
Of the five drives, one for File History, two for Storage Spaces, and the remaining two for alternating local backups (does robocopy still work?). USB C on the back allows transfer to a drive to be taken off site.
I like that drives remain accessible on their own and just one connector to my laptop hub. Not maintaining a NAS is a plus.
How terrible is this?
SHR works well in my experience on 4 drives. They use BTRFS on top of it (only the good bits, not the bits of btrfs that are suicidal to use). I've not had drive failures, but I have tested it by taking drives out (cue a lot of beeping and complaining but no problem rebuilding the array afterwards). It also scales up nicely; you just sequentially put bigger disks in, and the pool / volume simply grows.
I think I accidentally responded to the wrong thread. I was trying to respond to the thread where Bob.Brown claimed that 3-2-1 doesn't cover all the backup scenarios, and I got confused and thought the "smoking hole" scenario concept was a furtherance of that discussion.3-2-1 absolutely does cover the claimed scenario. I was responding specifically to quoted user's implication that Kevin setting the terms of the conversation by laying out 3-2-1 was just "nerds" "trying to gate keep what IS and what ISN'T."
Man, I was a big proponent of WHS back in the day; was absolutely devastated when they abandoned the product. I consequently continued to run it way too long after it went out of support. I don't know if it was my Gigabyte mobo or the WD drives I used, but I had more drive failures over time than in any other machine I've ever built. And every time time the system disk failed that was a pain in the butt.Cue my wistful yearning for the good ol' days of WHS (Windows Home Server) v1.
Automated, daily, incremental backups. Access to your choice of backups, all as a normal appearing Explorer window. Bare metal restore. File sharing, through easy network shares of Pictures, Music, Videos, etc. Drive Extender magic. And it integrated perfectly with Windows Media Center.
I've seen things you people wouldn't believe. A home server with 21 terabytes of storage in 2010. I watched Windows Phone tiles flipping in the dark near the Zune app. All those moments will be lost in time, like local-only accounts in Windows 12...time to switch.
ROFL, right, it's a conspiracy from Big Ferrite Donuts. Dude, again, just because you have not personally seen a thing does not make it not real.Let me ask this: Of all the external power supplies you've handled, how many have chokes on them? I'm looking at 13 power bricks in my (extended) vicinity and only two have them. Two, and one is a device from the 90s.
I've even seem USB cables with chokes on them. What do they know that that the thousands of other manufacturers and vendors don't? (Answer: They know that some people will buy anything.)
So either the rest of the industry has solved this, or they don't care about this supposedly widespread problem and yet still manage to stay in business and get great reviews that never mention how the electromagnetic radiation from the power cords has taken down civilization.
You know who loves this supposed problem? Ferrite donut manufacturers. They love it like undercoating companies love car dealers.
Sure, but that's not what I'm saying about SSDs. The storage medium SSDs use literally damages the memory storage location every time a new bit is written to it. That's not just wearing out with time, it's inherently destructive. With HDDs, however, some die sooner than others but it's based on general wear and tear due to entropy, not something which is inherently destructive in the performance of its job.You can believe me or not, but it doesn't change reality.
HDDs wear out from use. They absolutely die a guaranteed death through use.
Energy assisted hard drives burn out the heads over time. It's inherently destructive as they experience rapid heating and cooling. Older standard HDDs wear out via seek fatigue as well as a lot of other factors.Sure, but that's not what I'm saying about SSDs. The storage medium SSDs use literally damages the memory storage location every time a new bit is written to it. That's not just wearing out with time, it's inherently destructive. With HDDs, however, some die sooner than others but it's based on general wear and tear due to entropy, not something which is inherently destructive in the performance of its job.
Big ferrite donuts won't cure a baby monitor running on 2.4 GHz interfering with WiFi. That kind of interference has nothing to do with cables. Nothing you described really has anything to do with cables.ROFL, right, it's a conspiracy from Big Ferrite Donuts. Dude, again, just because you have not personally seen a thing does not make it not real.
I've seen this stuff happen devices more than once. It happens most when a device has multiple wires that are straight and lack shielding. I was first exposed to it as a thing when I supported Microsoft Hardware and one of the engineers designing the products showed me what could happen with a WiFi router places next to a Force Feedback Pro joystick. It's easy to see with the proper equipment but you don't need that to rule it out by simply changing a device's positioning. I've since experienced is a number of times over the years when moving a WiFi router a foot to one side or the other made an issue go away where placing it back in the original spot made it recur. This isn't something that's quite what I'd call common but it's also not rare, either.
Just recently I had a client who was convinced "her WiFi had been hacked". From her description of the problem, it sounded a little like a possibly failing WiFi router to me since it would intermittently just not work. She'd be connected but no data could be sent or received. She thought it was hacked because it coincided almost perfectly with her infant's nap schedule. After some discussion I decided to go out and look at the environment since something was clearly going on. Sure enough, she was turning on a baby monitor in the little one's nap room which was sitting next to the router. Turning the baby monitor on and off replicated the issue at will. Moving the monitor to a table on the other side of the room resolved it entirely.
The sorts of cables with that sort of shielding on them are made to be used in the relatively uncommon but real situations where it matters. Sure, it doesn't matter for the vast majority of cases but when it does, having one available is rather useful. Whether the specific hardware here needs one or not can't be determined forem the available evidence but the simple fact is they're absolutely useful in certain situations.
I agree this is true of system backups for companies that want to reduce downtime.
For an individual, there is also value in reliable long-term file storage: knowing that if there's a flood or fire or hack, you'll still have your photos / documents / writing. That doesn't need to include the OS, programs or settings.
This product seems to tackle the second, and although I'm not convinced it's a great-value option, a good solution could be useful to many people.
Two years ownership here, no complaints yet, though I've not had to rely on their customer support.I just can't trust Synology stuff. I've had a couple clients that had too many issues over the years with their Synology stuff and Synology's support was of little to no help. They've got neat stuff with great feature sets but their quality leaves something to be desired, IMO. Unless they've made significant changes in the last 5-ish years I just can't trust this sort of thing from them.
The point is about how folks don't recognmize RF interference being a thing. It's an example of RF interference that ordinary folks don't recognize as such despite it being pretty obvious once you know RF interference is indeed actually a thing.Big ferrite donuts won't cure a baby monitor running on 2.4 GHz interfering with WiFi. That kind of interference has nothing to do with cables. Nothing you described really has anything to do with cables.
Refuses to run on my hardware (which is run of the mill consumer stuff, not brand new, not ancient, like ~3 years old?). In fact it does something bonkers. If I have an unraid USB in the server when it boots it breaks the motherboard's ability to see USB keyboards and mice until I do a CMOS reset with the jumpers on the motherboard. I was utterly at a loss on that one. Tried it with four different thumb drives, from four different manufacturers, made 2-3 different ways each.Unraid?
From her description of the problem, it sounded a little like a possibly failing WiFi router to me since it would intermittently just not work. She'd be connected but no data could be sent or received. She thought it was hacked because it coincided almost perfectly with her infant's nap schedule. After some discussion I decided to go out and look at the environment since something was clearly going on. Sure enough, she was turning on a baby monitor in the little one's nap room which was sitting next to the router. Turning the baby monitor on and off replicated the issue at will. Moving the monitor to a table on the other side of the room resolved it entirely.
That looks nice. If I ever ditch windows storage spaces for my big dumb file server I'm definitely gonna look into whether they have a >=6-bay device that has SHR and allows either USB or eSATA so that post-setup I could just use it as a dumb barrel of data and not have to worry about long-term security support. I'm looking for a way to merge together a messy pile of disks into a 10-20TB pool and then not think about them too much for a decade if I'm lucky. I like these flexible striped parity setups because I'm cheap and I only need enough redundancy for it to not be a complete pain in the ass if one disk fails, because I back up vital stuff off this mess in other places/ways.Funnily enough, Synology offers their own solution for that on their larger NAS - SHR. https://kb.synology.com/en-us/DSM/tutorial/What_is_Synology_Hybrid_RAID_SHR
I've used it (though on just a 2 disk array, so not really testing a lot), and was able to restore when one drive failed after a lightning strike. YMMV, not a backup, just a variation on RAID (as Synology itself says at the end of that article), etc.
It's an example of RF interference that ordinary folks don't recognize as such despite it being pretty obvious once you know RF interference is indeed actually a thing.
Have you looked at how cheap new disks are? If you're turning "a pile" of old disks into a measly 10TB array, it'd probably be easier and cheaper to just buy a new disk or two. Last time I picked up 18TB drives, they were $200 a pop.That looks nice. If I ever ditch windows storage spaces for my big dumb file server I'm definitely gonna look into whether they have a >=6-bay device that has SHR and allows either USB or eSATA so that post-setup I could just use it as a dumb barrel of data and not have to worry about long-term security support. I'm looking for a way to merge together a messy pile of disks into a 10-20TB pool and then not think about them too much for a decade if I'm lucky. I like these flexible striped parity setups because I'm cheap and I only need enough redundancy for it to not be a complete pain in the ass if one disk fails, because I back up vital stuff off this mess in other places/ways.
There is actually something quite a bit like Time Machine built into Windows. They call it "File History" and it's a bit more faffing to set up (but nothing hard at all). It can work with either an external drive or a network location, and really works pretty well.How is this BeeStation different than a low end Synology NAS other than having a preinstalled drive and not making it easy to replace? I've got a QNAP NAS but all the features here sound like things that would be present on one of their NASes. Did they just add some setup wizards to a Single drive NAS to make it a little less scary? Are the Bee Apps actually different or just a simplified interface for their normal app?
I do wish Windows had something built in comparable to Time Machine. Being built in it's much easier to get people to actually use it and pointing my wife's new MacBook Air at the QNAP nas was pretty simple. Now I don't really have to worry about it. I've got backups for files for multiple versions and even a full system restore if I need it.
One thing I do wish the EU would look into is Apple iCloud. I'd love to be able to have the full iCloud backup with my QNAP nas or my MS one drive included in my Office 365 sub. Sure, Apple might not be able to guarantee the reliability of a third party back up but I'm okay with that. Failing that, I'd love a reliable way for automatic photo backups. I get why Apple wants to keep a tight lid on things running in the background on phones but it's pretty annoying to have to manually launch the QNAP app to be 100% sure to get photo back ups.
I've seen it once. A length of Cat3 was running alongside the AC power cord. I just could not get a consistent 10 Mbps transfer rate across that cable. Swapping the Cat3 cable didn't help. Once I re-routed the power cord, I was able to get 10 Mbps.As someone who has probably dealt with a number in the mid-hundreds of power supplies over my lifetime, not once have I ever encountered any hint of electrical interference caused by a power cord. And a good number of those little fuckers didn't have ferrite donuts on them. And even $9 power supplies from A* can be had with unencumbered power cords and they work. just. fine.
I suspect this is part of an industry mentality that never questions their own assumptions and lazily says, "This is how we've always done it."
(And I won't even get into a separate rant about having to use an external power block in the first place rather than incorporate it inside the enclosure so we can use a smooth end-to-end mains-voltage power cord for a total cost upgrade of about 13 cents.)
Used Synology in enterprise work and home use for years. So far, noo issues.Two years ownership here, no complaints yet, though I've not had to rely on their customer support.
Running along fluorescent ballasts can cause issues with 5 and 5e. Not sure on 6 as I have never run across that situation yet.zfs-snapshot-mgmt
I've seen it once. A length of Cat3 was running alongside the AC power cord. I just could not get a consistent 10 Mbps transfer rate across that cable. Swapping the Cat3 cable didn't help. Once I re-routed the power cord, I was able to get 10 Mbps.
I haven't had any problems with Cat5 or Cat6 cables though. I haven't played with Cat7 or Cat8 yet.
Given that modern cables are shielded themselves, it's rarely an issue.I've seen it once. A length of Cat3 was running alongside the AC power cord. I just could not get a consistent 10 Mbps transfer rate across that cable. Swapping the Cat3 cable didn't help. Once I re-routed the power cord, I was able to get 10 Mbps.
I haven't had any problems with Cat5 or Cat6 cables though. I haven't played with Cat7 or Cat8 yet.
Given that modern cables are shielded themselves, it's rarely an issue.
I think it can cover the smoking hole scenario with caveats.Because something doesn't cover the "smoking hole" scenario doesn't make it not a backup. However, one SHOULD plan for the smoking hole, I agree, if you really don't want to lose data.
But I don't understand why 3-2-1 isn't a backup, as is claimed? 3-2-1 covers the smoking hole because one of the copies is offsite, on a server, one copy is onsite, on a secondary device, and one copy is on your local hard drive (or maybe one copy is on a second external device like a USB stick, BD-ROM or whatever.
How is that not good coverage for the smoking hole?