Neoseeker : News : A big improvement in small storage.
Sort by date: ascending descending
0 thumbs!
^
Shadow of Death Sep 6, 07
Cool...this will suffice until other technologies eventually replace this...
0 thumbs!
^
Smoke Sep 7, 07
Wow just think of all the...

" and to beat someone to the punch, I'll be the first to mention the XXX side of the Internet."


...crap =(
0 thumbs!
^
The Slayer Sep 11, 07
You could download the whole internet to one of those drives in a few years!!!
0 thumbs!
^
Cillchaoi Sep 11, 07
Micah, I just read this article of yours and I must say that I like the sound of the upcoming technology. However, I want to point out some details regarding the idea that you suggested, specifically making a RAID 0 array of eight 2TB drives. That would result in 16TB of storage but there would be no fault tolerance. This means that if a single drive fails, all the data across all the drives would be irrecoverable. This is the reason that RAID 0 is not commonly used. The best configurations would be RAID 3 (data striping with dedicated parity and parallel access), RAID 4 (data striping with dedicated parity and independent access), or RAID 5 (data striping with distributed parity) in that they create 14TB arrays with 13TB of usable storage with fault tolerance, fault tolerance being the main purpose of using RAID arrays. RAID 5 is the most common of these three options.

If RAID 0 is used, I strongly recommend having a backup system that is used on a frequent basis, such as a nightly backup routine (full backup once a week and incremental backups all nights at a minimum) or a live backup system such as the Seagate Mirra Personal Server (Seagate bought out Mirra nearly two years ago), especially when the technology is initially marketed. As we all know, new technologies tend to be prone to failure. Thus, to protect the data against drive failure in a RAID 0 configuration, a well-administered backup routine would be necessary.
0 thumbs!
^
MicahWrites Sep 11, 07
Well Cillchaoi, I personally don't like RAID configurations. However, those people who do dabble with RAID will commonly start with A RAID 0 just to boast of a multi-terabyte system.

Yes, striping and mirroring is important for data farms and large cooperations who want to protect their data, but most home users don't expand further into levels of mirrored RAID 1, 0+1 and or striped RAID 5 with some RAID 1 for mirroring.

Fortunatlly, HDD MTBF (Mean Time Between Failure) rates have never been lower, so data loss is a once in a blue moon occurrence.

Just having a 16GB drive C: sounds cool!
0 thumbs!
^
MicahWrites Sep 11, 07
Oh, and the stupid 1024 bye versus 1000 byte formatting difference, we can't forget that unavoidable loss.
0 thumbs!
^
Cillchaoi Sep 11, 07
Micah,

Being in the business of computer sales and repair (as I have owned my business since 1989), I have seen an increasing number of home users who have used RAID 0, RAID 0+1, and other configurations (even RAID 5) and have had the unfortunate duty to inform these people that when one hard drive died in their array (if lower than RAID 3), everything was lost. This is why I made the point in my post above: I wished people to be aware of the problems with the aforementioned RAID configurations since you brought up RAID 0 in your commentary on the emerging storage technology.

As for Mean Time Between Failures, you actually want a high number, not low, but I understood what you meant. I shall point out one other thing, which is that in the past two years, I have bought two brand new hard drives (one Maxtor IDE and one Seagate SATA2) that died within 24 hours of installation. Admittedly, immediate failure is not the norm but it does help to demonstrate that failures are not something that can be pooh-poohed away either. Thus, as I said, backups are always a good idea, whether one is a home user or a corporate businessman.

About the comment regarding 1000 versus 1024, that is not a fault of the format but, rather, the way that hard drive manufacturers advertise the size of the drives. If one reads the (limited) documentation that comes with the hard drive or that is available at the manufacturer's website, one will find that the literature says that they measure a megabyte to be 1 million bytes and a gigabyte to be 1 billion bytes. The computer measures it out correctly to be 1,048,576 (2 to the 20th power) for 1MB and 1,073,741,824 (2 to the 30th power) for 1GB. Thus, a 500GB hard drive, as advertised, would actually be 465GB according to the computer.
0 thumbs!
^
MicahWrites Sep 11, 07
Well, because you own and operate a computer store it is normal for you to see failed components on a regular basis. I may not see many people with broken arms but a doctor at some hospital will.

And no need to quibble over the semantics of bit counting in the industry and in the OS. It's like monitors with an advertised 21" screen which in reality have a 19.7" viewable area.

Hmmm, as I recall some people had tried to sue manufacturers for that under the guise of mis-leading and or false advertising.

Still, I've had some HDD's running for 4-6 years (always on, either seeking or sleeping) with not a single cluster lost. I did have an IBM DeathStar HDD grind its head into the platter after 90 days. But on average, HDD's do have a pretty long life span now.
0 thumbs!
^
Cillchaoi Sep 11, 07
Yes, on average, that is true. My point, though, which remains true and has been what I have been stating all along, is that backups are recommended, whether one is using standalone drives or drives in a RAID 0, 1, or 0+1 configuration. RAID levels 3, 4, and 5 have fault tolerance built in so that the failed drive can be rebuilt on the fly once it has been replaced. Therefore, RAID levels 3, 4, and 5 may be backed up for redundancy but is not as necessary as the other mentioned configurations.

Commenting on the issue of viewable as opposed to full physical CRT size and the attempted lawsuits about that difference has no bearing here. Once again, you made a comment that contained information that could mislead people: your comment seemed to imply that people were "losing" storage space because of the format applied by the OS when that was not accurate. I was only bringing the accurate information to light.

Further, you are correct that I see more failed components than a typical end-user since I am in the technology sales and service business but that does not mean that failures do not happen and that they are not a matter of concern for the wise user. As any technology professional can tell you, backups are recommended for everyone to protect against data loss. The issue is not if it will happen but, instead, when. If one uses computer technology long enough, a hard drive failure will occur. The laws of averages will catch up to the user eventually, whether it be sooner or later.

Mean Time Between Failures is a measurement of that average as tested in an accelerated test environment. The mathematical definition of mean is all values added together divided by the number of values in the list. Nowadays, MTBF is measured in hundreds of thousands of operational hours but that means that one test could have a result of 2000 hours (about 83 days) and another have a result of 198,000 hours (22 years 220 days) resulting in a mean time of 100,000 hours (11 years 152 days).

Thus, as drive failure is unpredictable (though SMART technology helps to alleviate that), backups are recommended to protect data (documents, pictures, digital home videos, etc.) that one does not wish to lose.
0 thumbs!
^
iamjoe56 Sep 11, 07
Micah, their is something you should know. Fortune favors the prepaired. Now if you run a Hard drive set up with NO failsafes, you WILL have a failure, and lose everything. Also, newtons law, Evrything in the universe tends towards entropy. IT is true you know.

Basically what you have said so far is that nothing will happen if you are careful. You can be a carful driver, and still have an accident. Nothing is ever certain, you keep treating matters as if they are.

I believe at this point it is time for you to sit back, read what Cillhaoi has said, and reconsider. This is one issue you are not going to be right on, trust me.

As for the Bit counting remark, I have got to refute that, even I know better. If you do not state technical issues like that correctly, you may cause someone who does not know this issues well, astray.
0 thumbs!
^
MicahWrites Sep 11, 07
iamjoe56,

I never said that HDD's don't fail; most last a very long time and a few will drop clusters in mere minutes. Furthermore, I had never denounced the option for people to use striped RAID arrays, I had simply written that 'imagine 2TB drives? Try 8 in a RAID0 array for 16TB worth of whatever you want.'

It was Cillhaoi who decided to elaborate and split hairs over what was simply a colorful statement. Yes, people should protect their data.

As for the bit counting remark; I was responding in context to the previous post by Cillhaoi who had remarked that you will have reduced capacity with any form of RAID redundancy, as you will with the 1000 vs. 1024 bit count issue. Such a simple comment with no expository explanation cannot lead astray.

If I tell you that 'The sky is blue.', is that enough to endow you with the new knowledge of the refractive index of Earths atmosphere or the wavelengths of light with the electromagnetic spectrum.

Context young man, learn to remain within the context of a conversation.
0 thumbs!
^
Cillchaoi Sep 12, 07
Micah,

If you were not advocating the idea of a RAID 0 array, why mention it at all? You could have made the same claim of eight 2TB drives giving 16TB of storage without mentioning RAID 0 (though it would not have sounded so “cool” since it would not have used technological buzzwords). By specifically mentioning RAID 0, you seemed (to a typical end-user) to recommend its use. Many end-users know the term RAID has to do with data integrity but do not understand the complexities of it, such as the different possible configurations. Thus, a typical end-user could believe, based on your perceived recommendation, that buying lots of drives and putting them into a RAID 0 array could give them enhanced data storage and data integrity without the need for backups, killing two birds with one stone. That is not the case and is to what I commented and why I gave a full explanation of RAID 0 and its downfalls in response.

You are correct that Mean Time Between Failure measurements nowadays are much higher than they were back in the days of MFM, RLL, and ESDI. Back in the late 1980s and early 1990s when these technologies were most prevalent, MTBFs were typically about 10,000 hours (about 1 year 52 days). Nowadays, as I pointed out previously, MTBFs are measured to ten times that number. MTBF is still only a statistical average and is not to be taken as a prediction of actual time of failure for any individual drive. Backups are still recommended by all professionals because of this fact.

As for the comment about “splitting hairs,” I shall say that I am not splitting hairs but, rather, providing more information than you did and correcting some errors that you made in your article as well as your comments thereafter. In any professional publication, one must write as if someone who knows more about the subject matter will read the material because, invariably, someone more knowledgeable will read it. This is what seems to be happening here (though I cannot say the definitively since I do not know your background; I know only mine). I would advise you to make sure that you have all the information correct prior to publishing your articles.

The “1000 v. 1024 bit [it should be byte rather than bit] count issue,” as you said, in your one-line comment, was a “formatting difference.” That seemed to imply that you were under the mistaken impression that it was an issue of the operating system rather than an issue of human usage of the term as opposed to the true definition of the term, especially when you concluded it with “we can’t forget that unavoidable loss.” You did not mention anything about RAID levels in that one-liner, so trying to say now that you were intending to make a comment on RAID is unbelievable. Once again, your words could make end-users who have a lack of knowledge believe something that is not true, an error that I discussed and corrected with factual information.

Yes, RAID levels 3, 4, and 5 will lead to a loss of 3TB of storage in an array of eight 2TB drives but that 3TB of loss gives reliability and redundancy. A failed drive can be rebuilt on the fly by the array once a new drive is installed. Further, in RAID arrays such as those I mentioned, a hard drive failure does not incapacitate the computer in which they are installed. Instead, the machine stays up and running as if nothing is wrong but, instead, displays a message stating that a certain drive in the array has failed. With RAID 0, which you advocated in the original article, that is not the case. Instead, all the data would be lost unless one had wisely made a backup of all data. Further, this would cause a greater loss of time because all programs and the OS would have to be reinstalled before the restore was able to be performed (unless someone just happened to have another 16TB of storage just lying around for a total image backup by the means of Norton Ghost or something similar).

Now as for your analogy about “the sky is blue,” once again you make an error in trying to draw your analogies, as I pointed out in another article discussed a few days ago. If you wanted to draw a proper analogy for the discussion of formatting difference, it would have been more like saying, “the sky is blue because air is blue.” That is not accurate but, to the typical layman, that would appear to be the case. If this were a scientific discussion, I could delve into this and explain the facts but that is not what applies here.

Finally, you have said that Joe should keep his comments in context. Reviewing his commentary, I see nothing that is out of context in regard to what he said in response to this conversation, especially to what you have said yourself.

concluded...
0 thumbs!
^
Cillchaoi Sep 12, 07
Micah, I think that, generally speaking, you have a good head on your shoulders but it seems apparent to me that you need to be more careful about the information that you present as well as the manner in which you present it. From what I can tell, you have been tasked to write articles about technology. The articles themselves should present the facts, not your opinion. I would suggest that you save the professional opinions for the follow-up comments or for a weblog (blog is the commonly used term for those who do not know the full term). The personal opinions (such as “the stupid 1024 bye [sic] versus 1000 byte…”) should be withheld. To present personal opinions such as those shows a lack of professionalism, something that detracts from NeoSeeker’s public image as you are a representative of them in the public eye.
0 thumbs!
^
iamjoe56 Sep 12, 07
"you will have reduced capacity with any form of RAID redundancy".

That is presicely why my comment was well within context, and perfectly fair. It seems it is you who do not understand the concept of "context". Also, the sky has presicly no bearing on this conversation, thusly, that part was totally OUT of context. Second. You NOT mentioning the fact that a RAID configuration with any type of redundancy would reduce capacity, would lead some people astray. Very true not knowing your storage amount would not cause major problems, but such habbits like putting forth misleading information could lead to a much more serious blunder later on. My Main point is, putting forth untrue facts, is very misleading and dangerous (in a few extreme cases). And Secondly, do NOT try to flaunt your intelectual "might" over me. I am just as smart as you are, brain cell for brain cell. If you refuse to treat me as an intelectual equal. Then I will treat you in the same fashion.

That did not work last time, and will not work this time. Simply learn to accept GOOD advice, and try to verify your information before placing into one of your articals.
^
Sponsored
Sort by date: ascending descending
- This news story is archived and is closed to new comments now -
Newsletter
Latest News
Latest Inhouse
Deepcool Maelstrom 240 Liquid Cooler Review XFX R7 250E Core Edition Review Scythe Mugen Max Cooler Review XTracGear Mouse Surfaces Review Diamond Dual Band Wireless 802.11n Range Extender Review Noctua iPPC and redux Fan Roundup
(0.2557/d/web3)