SoylentNews is people
SoylentNews
Submitted via IRC for Bytram
There's one crucial way tape still trounces SSDs and hard drives when it comes to storage...
Tapes will make sense to those born in the early 80s but magnetic ribbons have long been replaced by shiny disks, silicon chips and cloud-based storage for whoever wants to store data.
But don't discount them yet. Even if LTO-8 tapes are now in stock, you can buy cheap LTO-7, reformat them to M8 and get 9TB of native storage (22.5TB compressed). You can grab one (HPE LTO 7 Tape with Barium Ferrite (BaFe) C7977A) for just under $59.
With an uncompressed capacity of 9TB, it translates into a per TB cost of $6.55, about 12x less than the cheapest SSD on the market and 1/4 the price of the 12TB Seagate Exos X14, currently the most affordable hard disk drive on the market on a per TB basis.
In other words, if you want a LOT of capacity, then tape is the obvious answer (although truth be said, you also need to factor in the cost of the drive).
But there's something else that tape offers that no other storage medium currently offers and that's on-the-fly, transparent compression which can go up to 2.5:1 and works best on text files (rather than multimedia which is already heavily compressed).
As for transfer speeds, they can reach 300MBps (that's 1.08TB per hour) which is plenty fast, just a tad slower than the just-reviewed PNY Pro Elite which tops 375MBps.
Tape that!
(Score: 5, Insightful) by Immerman on Wednesday September 25 2019, @01:34PM (19 children)
Excellent compression... on text files, which are already a tiny portion of most people's data collection.
Arguably good for businesses that may have lots and lots of text-based data, except that these days most of that data is probably in one of MS Office's compressed container formats which will see little if any additional compression.
And of course there's compression's biggest benefit for archived data: Greatly reduced data integrity! Because losing a few bytes of data in the wrong place in a file should guarantee that the file becomes completely unreadable!
(Score: 4, Interesting) by sgleysti on Wednesday September 25 2019, @01:50PM (10 children)
Fun fact: Office Open XML containers are just ZIP files with certain requirements on what settings are and are not allowed. I wanted to send an Excel workbook to someone, but it was over their email receive limit so I unzipped it and rezipped it with maximum settings in 7-zip. Voila, it was now under their email receive limit. My guess is that Office doesn't use the maximum compression level because it would make saving files too slow.
(Score: 2) by FatPhil on Wednesday September 25 2019, @02:17PM (9 children)
Great minds discuss ideas; average minds discuss events; small minds discuss people; the smallest discuss themselves
(Score: 3, Informative) by DannyB on Wednesday September 25 2019, @02:32PM (8 children)
I can confirm Java JAR files are Zip files. But pay attention when re-zipping, I seem to recall having trouble with some highly optimizing zip compressors.
Also: LibreOffice documents are Zip files. And they were zip before Microsoft's Office Open XML format mentioned above.
Other fun fact: You can Export (not Save) a LibreOffice document as a Hybrid PDF. This means it is not only a PDF, but has the LibreOffice format (a zip) embedded into the PDF. Thus it can be read like a PDF. But it can also be re-opened in LibreOffice, retrieving the original office document in its fully editable form.
The lower I set my standards the more accomplishments I have.
(Score: 0) by Anonymous Coward on Wednesday September 25 2019, @02:39PM
n/t
(Score: 3, Informative) by shrewdsheep on Wednesday September 25 2019, @02:41PM (6 children)
That is natural, as MS Office Open XML format is an embrace, extend, extinguish reaction of MS to the Open Document Format (ODF) as used by LibreOffice. Also, MS is heavily represented in the ODF bodies, so that is that.
(Score: 5, Informative) by DannyB on Wednesday September 25 2019, @02:49PM (5 children)
This was disgust on Groklaw quite a bit back in the day. Microsoft first of all called it Office Open XML so that it would be shortened to OO XML. Thus deliberately causing confusion with OpenOffice.org (later LibreOffice) XML. When you see OO XML you would think OpenOffice.org XML.
Microsoft made sure that its implementation of MS Office wasn't really quite fully compatible with the spec for OO XML. Naturally MS Office was the primary and perhaps only implementation of OO XML. This was intended to deliberately frustrate attempts of OpenOffice.org to interoperate properly.
The entire OO XML happened because OpenOffice.org proponents were successfully getting organizations, and even governments to require office documents to be in an open format to protect the long term readability of government documents. These docs might need to be readable for a hundred years. Do you think MS Word 2100 is going to be able to read today's Word doc? So OO XML technically satisfied that requirement. Reluctantly. But Microsoft still worked to frustrate true interoperability.
The lower I set my standards the more accomplishments I have.
(Score: 3, Funny) by Runaway1956 on Wednesday September 25 2019, @03:19PM (1 child)
I read that, and thought, "DannyB spelled it wrong, and didn't say what he meant." Typos do that.
After reading the post, I looked back, and realized the typo was probably intentional. If not, is it one of those Freudian slip things? :^)
(Score: 2) by DannyB on Wednesday September 25 2019, @03:30PM
Intentional.
The lower I set my standards the more accomplishments I have.
(Score: 2) by shortscreen on Wednesday September 25 2019, @06:53PM (2 children)
You left out the part of the story where MS went through and stacked the ISO committees with their buddies so they could get enough votes to have their format approved as a standard.
(Score: 2) by DannyB on Wednesday September 25 2019, @07:29PM
There was A LOT more to it at that time than I mentioned.
During this whole OOXML vs OpenOffice.org fiasco was when I first started hearing the name of a state governor named Mitt Romney. He played a part in the state's selection process, and some people smearing the name of a state IT professional recommending the state use open source software.
The lower I set my standards the more accomplishments I have.
(Score: 2) by chromas on Wednesday September 25 2019, @10:09PM
Same thing happened with Debian and systemd.
(Score: 2) by The Mighty Buzzard on Wednesday September 25 2019, @01:58PM (2 children)
Tape isn't for "most people" though. It'd be nice for us if we were a bit bigger, for instance, since pretty much everything we'd want to save is text and we manage quite a bit of it. Either way, it's still a really, really good price on archival storage even without compression and lasts a lot longer than optical media if you don't buy garbage.
My rights don't end where your fear begins.
(Score: 0) by Anonymous Coward on Thursday September 26 2019, @04:37AM (1 child)
Do you not want to save your log files? Surely they would be in a much more usable binary format!
(Score: 2) by The Mighty Buzzard on Friday September 27 2019, @11:12AM
They would indeed, because gzipped text files count as binary.
My rights don't end where your fear begins.
(Score: 3, Informative) by FatPhil on Wednesday September 25 2019, @02:13PM
> But there's something else that tape offers that no other storage medium currently offers and that's on-the-fly, transparent compression which can go up to 2.5:1 and works best on text files (rather than multimedia which is already heavily compressed).
Every other storage medium still in existence offers that. It's a service provided either:
1) transparently by the OS through the choice of a file system that supports compression; or
2) transparently by the backup software that you use to backup your files (which might even be something as simple as 'tar').
And, guess what - by pulling the compression into a higher layer, you can also customise things like your own desired levels of error-detecting/correcting codes, or even more high-tech or domain-specialised compression algorithms. Saying it's better because it's hard coded into the drive itself (which is the only way the above claim makes any sense at all) is praising an antifeature.
*And* what you said about the meaningless compression ratio claims, which are even more exagerated than Stacker, or whatever it was called, were claiming back in the early 90s. They're probably relying on the youngsters not remembering those days, and remembering how pointless those claims were (for the reasons you mention).
Great minds discuss ideas; average minds discuss events; small minds discuss people; the smallest discuss themselves
(Score: 5, Interesting) by JoeMerchant on Wednesday September 25 2019, @03:17PM (3 children)
We did a tape based system in the early 1990s - FedExing tapes from all over the country to a central storage site. I forget the particulars, it was "state of the art high performance tape backup storage" for the time, and my experience with similar systems before and after was... similar.
In this capacity, the tapes were "tested" in a way they normally are not for things like on-site backups. Every bit of data had to be successfully restored from every tape sent through the mails. Due to the absolutely horrendous error rates we were experiencing (like 1/10 tapes made had at least one unrecoverable error), the common practice became: when making a tape backup, make 3, mail 2 to the central office and keep the data on the hard drives until the central office confirms successful recovery from at least one of the two tapes.
There were 5 sets of tapes flying into the central office per week, so about twice a year they'd have a double failure and have to call on the remote site to make more backups... this pattern continued for over 2 years.
When you factor in all that, the 1/12th cost per byte "stored" isn't so good.
🌻🌻 [google.com]
(Score: 0) by Anonymous Coward on Thursday September 26 2019, @12:07AM
At a prior employer a single tape set for a full was 250 tapes across multiple tape libraries. We frequently had a tape with read errors in a set when we tried to restore. But, what you describe indicates something seriously wrong. Our off-site tapes were stored by Iron Mountain in environmentally controlled conditions, and we kept track of tape insertions, and after a threshold (can't remember the number) the tape was disposed of.
A solution to these occasional errors is to use erasure coding i.e., tape raid. A bunch of backup software supports this. And, pretty much everything short of mt keeps track of insertions for you.
(Score: 2) by Immerman on Thursday September 26 2019, @01:42AM (1 child)
Did you send the tape in insulated packages? As I recall, tape can be rather vulnerable to the sort of extreme temperatures that mail is often subjected to, not to mention humidity. I'm pretty sure the big players tend to use climate controlled storage/retrieval facilities for tape, or at the very least a nice out of the way closet
(Score: 2) by JoeMerchant on Thursday September 26 2019, @02:12AM
This was the early 1990s, if you consider cardboard and some packing paper insulation, then yes.
Thing was, the error rate was high (maybe not 1/10) even at the local sites without the mailing step. The damn things were flaky - new from the package, backed up one day and restored the next. We weren't asking them to do anything unusual, besides work most of the time, which they didn't seem capable of. That study enrolled something like 500 subjects for about 8 weeks a piece, and I think they made it through without losing more than a couple of weeks of data permanently, so... good enough, but a lot of hassle to still lose some data.
I remember similar error rates when using tape for on-site backups both before and after the study - they worked... most of the time, and you needed them so rarely that most of the time was good enough, except when it wasn't.
🌻🌻 [google.com]