Well, you do have to refresh your backups on any medium, including HDDs. Taiyo Yuden discs have been very reliable for me for about 15 years now. I wouldn't guarantee that a HDD sitting on a shelf would start up reliably at that age.
Not true. Synology uses a Linux distro, and there's a dedicated ransomware version (SynoLocker) aimed at them.
Not against APTs--they can reflash your hard drive and/ or BIOS, and operate from there. The ATA command set includes "load microcode", so it doesn't even need that much hacking.
And even a live CD doesn't help once your disc has been encrypted--it just makes it impossible to recover.
Cheers
Phil Hobbs
--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics
160 North State Road #203
Briarcliff Manor NY 10510
hobbs at electrooptical dot net
http://electrooptical.net
In short, yes. The Phi cores themselves are mostly x86-compatible, but the differences are at the upper level: they are shipped as a PCI express card with its own OS (ssh-able). At least the ones I've seen.
OTOH, they support OpenCL wery well, so any GPU-aware program would also run smoothly on the Phis.
I got a Dell 5810 workstation earlier this year with a Xeon CPU, ECC memory and SSD. It's very quiet (I actually had to get used to the absence of noise in my office) and it has been rock solid even with very memory & CPU intensive tasks (routing FPGA designs and compiling large software projects).
The SolidWorks viewer shows all the surfaces nicely rendered and colored, and does sections so I can see inside things. Things spin about as fast as you could spin the real thing in your hand, on my old HP. I don't run the full SolidWorks, just the viewer.
This spins around with no visible delay:
formatting link
I couldn't spin the real thing as fast: it's big and heavy!
I would, if it were actually better. Scrambling the UI doesn't make it better, it just makes it annoying.
Windows 10 looks to be yet another disorganized Apple clone, of OS-X this time. X=10, get it? Steve Jobs was Microsoft's best architect.
--
John Larkin Highland Technology, Inc
lunatic fringe electronics
jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com
Yes, I understand. Its a different thing, entirely, to do full ray tracing, light sources, etc. I.e., I build models that I can "take photos of" and not know that there was no "camera" involved!
Exactly. But users only *see* the UI. Joe Average User couldn't describe Windows (any version) in terms other than "a graphical user interface with lots of WINDOWS"
On Mon, 03 Aug 2015 11:09:14 -0400, Phil Hobbs Gave us:
mSATA drive is an SSD. A spinning, platter based drive is a tiny clean room and would be no different now or in 20 years. Far more reliably than an optical impingement (it isn't even a true 'burn"). Hint, hint.
Using SSDs for long term backup is an option for trust fund babies, or those with very little to back up.
Taiyo Yuden CDRs and DVD+Rs are very reliable.
Cheers
Phil Hobbs
--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics
160 North State Road #203
Briarcliff Manor NY 10510
hobbs at electrooptical dot net
http://electrooptical.net
On Mon, 03 Aug 2015 18:09:36 -0400, Phil Hobbs Gave us:
The likelihood that you'll lose your data from a relaxation of the original write session (hint, it is NOT "a burn") is far higher than some stray gamma particle causing an SSD to lose what was written to it.
That is one 'fund' I'll trust forever by comparison, baby
Optical discs which were stamped by a glass master at a factory are reliable. The crap consumers "burn" are absolutely not (burned or reliable). The "pits" are mere slight impingements in a plastic layer made at a very high speed which relax over time or with elevated temperatures.
Which is why you must not shelve hard drives. I choose to have all my data online on multiple hard drives. That way I have health monitoring of the media. I found out I lost little bits of data by storing it on CDs or hard drives because I misplaced them or never cared to restore when changing computers. Making backups is one thing but organising the media in a sensible way is a cumbersome task.
You don't want to have them all mounted to a filesystem, though, because ransomware is spreading like a plague.
Cheers
Phil Hobbs
--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics
160 North State Road #203
Briarcliff Manor NY 10510
hobbs at electrooptical dot net
http://electrooptical.net
Unless you're actively *checking* the media contents while they're spinning, you still have no idea whether or not any particular file can be retrieved when you need it.
Don't bother "organizing"! I'm moving all my "bulk" archives onto external drives (so they can be easily/quickly moved without having to worry about building a "compatible" machine to read them WHEN something goes south) and loosely "mirroring" them -- with the mirrors being is convenient.
Then, logging the contents of each rive in a relational database -- along with applicable metadata (e.g., MD5's of each file, size, etc.). So, I can query the database *without* having any drives spinning -- to locate a file of interest.
Once located, I can spin up the necessary drive(s) and retrieve the file(s).
At the same time, a job systematically walks each filesystem WHEN the drive is on-line and reverifies the MD5's of each file against the MD5 stored in the database. So, if a file is in jeopardy (yields read errors or a bad MD5), I know *then* to hunt down the mirror copy and create a new backup.
Allowing the "verification" job to be interrupted (e.g., when the drive is spun down) and *resumed* (in place) means I don't have to leave a box up for hours while I verify its entire contents.
[At the same time, logging that task's progress (in the database) lets me know when portions of each medium have been "ignored" for too long...]
On Tue, 04 Aug 2015 02:52:22 +0200, "N. Coesel" Gave us:
I have several multi-terabyte drives laying here. Using one for films and music, and one for archiving business data does not pose an issue with my fail mind. (roll eyes again)
On Mon, 03 Aug 2015 19:37:25 -0700, Don Y Gave us:
Horseshit. If it was a reliable drive when you archived the data, and you make dual partitions even on the same drive and manually (not a mirror set up)redundify the stored data, you will not experience a problem.
You can access the data via a variety of protocols that are less accessible than a nominally mounted filesystem (assume the host that has them mounted is "secure"). E.g., FTP, RSYNC, HTTP, etc.
It's not just ransomware. When a disk is spinning and mounted, its contents can be altered (when its on a shelf, they can't -- shy of an EMP).
So, if the OS acts up (or the firmware on the drive), your data is at risk.
One reason I don't like COTS RAID arrays is there are still "shared components" involved for "both" copies of your data: power supplies, CPU's, firmware, memory, etc.
Decades ago, I had a ~dozen (bare) SCSI drives that I kept as a "cold" archive; two copies of everything. When I needed a file from the archive, I'd install a drive in an external SCSI enclosure, cable to the system in question and then mount(8) the filesystem present on the drive. Copy the file(s) I needed, then spin down the external drive and put it back on the shelf.
One day, I installed one, mounted it -- and watched it FAIL in a big way. My initial assumption was that the *drive* had failed: "(sigh) $1K down the drain!" So, I pulled the "bad" drive and mounted the mirror copy. And watched it suffer the same fate! "Hmmm... what are the chances of TWO drives failing within minutes of each other??"
Turns out, the OS had an "issue" in the driver for the SCSI HBA with that particular make/model drive -- something that didn't quite work right (these are formally called "QUIRKS" by the code). Until the issue had been identified, any drives of that type would suffer the same fate!
I.e., had both been mounted at the same time, both would have died at the same time as the problem lie in the OS/driver/HBA/disk combination. Had all ~dozen of them been spinning in a shelf, I'd be looking at an order of magnitude higher loss!
In this case, the problem manifested quickly and in a very obvious way; had the QUIRK been more subtle, it could have been less obvious and not discovered until a later date -- AFTER I had assumed that the retrieved file was intact!
This taught me the value of having backups on other forms of media (I recovered the contents of the drive and its mirror from a third copy on MO media) *and* the value of having drives with R/O strapping options! (i.e., try as it may, the OS *can't* alter the contents!)
Things break (aka "Shit Happens") so expecting *anything* and EVERYTHING to work reliably at all times is a Rx for disappointment!
They are provided you keep them cool, dry and in the dark. I am always a little nervous of DVDs for high value stuff - I prefer to keep important files on at least two independent media types.
It is scary seeing what happens to ones used as coasters left outside in the sunshine. Even a sunny windowsill will degrade lesser brands (especially if they forgot the UV protection to save money).
There is an old-school barrier to malware and ransomware - tape backup.
The reason tape is resistant is because it does not have a file system on it, and cannot be accessed randomly.
So one recovery dodge if one has a infected disk is to back all desired files up onto tape, reformat the disk, install a brand new OS image, and then restore the files.
This will not help if the files were encrypted by ransomware.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.