Hard disks, solid state drives

HDD, SSDForum → Discussion of the article: " Seagate Winchesters from Barracuda 7200. 11 to BarraCuda "

Discussion of the article: " Seagate Winchesters from Barracuda 7200. 11 to BarraCuda "

Pages: 1 2
24.04.2017 14:47:00
I propose to discuss the Seagate HDD from Barracuda 7200. 11 to BarraCuda
Thank you, interesting comparison
24.04.2017 16:51:00
So, the old working hard drive is changed to a new one. . . In general, there is no need.
Actually also applies to the RAM.
24.04.2017 17:43:00

Only here ST31500341AS to shove and into a modern computer can be without problems, and with the performance of 2008 there may already be some technical difficulties
24.04.2017 18:25:00

you in this line have missed something missing:
quote:
is now also does not interfere - just emerged SSD and necessary tasks such "load" rather than hard drives.
24.04.2017 18:50:00
Quote from the article
quote:
The top as chemotherapy even faster than today's younger BARRACUDA, so that instead of the first to buy the latter can mainly for repair
If you really want faster - perhaps it makes sense to look more expensive Constellation or At least VX " S SV35?
Although they are much more expensive.

By the way, is there a chance for a comparative review of hard disks of the same volume of different classes of the same manufacturer - for example, 4 terabytes of Barracuda, VX / Surveillance and Constellation?
(EMNIP, the last (NM) in the iXBT test never was, except for the helium 10 terabyte, which is strongly non-acidic.)
24.04.2017 19:02:00

Do not miss
There's still before point proposition is


quote:
probably makes sense to look more expensive Constellation
And why did you decide that the younger [the same capacity] will be faster?
The recall of the ST3000NC002 was, to put it mildly, no faster than ST3000DM001 in terms of desktop application at one time
Rather, even on the contrary
24.04.2017 19:13:00
quote:
:

quote:
probably makes sense to look more expensive Constellation
And why did you decide that the younger [the same capacity] will be faster?
And I do not know. Therefore, I ask about the comparative review of the same volume of different classes of the same manufacturer .


P. S. The main Bonus bonus I already know is reliability. But it would be interesting to look at the comparison of other parameters with more budgetary brethren.
24.04.2017 20:47:00
I have a screw 10 year old Samsung on 750.. The last year 3 is not enough speed, the games began to weigh 50+ gigs and load 4+ gigs per operative at a time, compared with new screws, a loss in the speed of 2 times. And do not let if the file swap calls, friezes omg, The benefit of the recently expanded the operative and until the swap does not climb the game. The next upgrade - ssd, specifically for the system and the current cool game, and all the other garbage can be left on the Samsung, all Blizzard games and music with movies and 10 years ago it was normal with it. It's a pity that even the discs of the Samsung do not release, the Gadish сигейты bought the production in order to produce its trash.
24.04.2017 20:52:00
The error there crept in. 1AS - five pancakes, and not four. ASCII_05_04B8E.jpg
The penultimate figure says this.

It makes no sense to compare the models of the same factory family, if they are part of a discrete or corporate group.
These are absolutely identical discs with minor changes in the passport settings
The spread of the parameters of one model, one family, can give a spread in the speed of up to 20%.
old model designation clearly belonged to the same family of factory
341AS Brinks
542AS Hepburn
651AS MantaRay

But with the new designations may already be miracles, the model has become hard to determine where to use which is intended.
And not what's in it.
The factory family in one model can be different or different modifications of one family from completely different generations of components.
The same model, released after a few months, is completely different and not compatible with the first release either in detail or in parameters.
While the technological utility does not look, and you will not know what it is and how many real pancakes there are, what density and what organization.
25.04.2017 9:32:00
TB per kilowatt in servers is very important, well, the physical dimensions of this whole farm are natural.
25.04.2017 9:40:00
The most disappointing thing is not so much the slow growth of performance (in principle, there is SSD for fast loading, and HDD for long-term storage of large volumes of information), how much the reliability of modern HDDs decreases (although, on the contrary, they could find themselves against the HDD As a long-term and capacious storage).
Bought a new disk shelf. For six months of operation a bunch of dead server WD (change under warranty). That is, the warranty will end - buy a new shelf or go broke for a constant replacement of flying inexpensive SAS " Disks.
25.04.2017 12:06:00

If you fly, then something is wrong with the conditions.
Almost all discs themselves have a failure caused by the purely mechanical nature of the disks at the level of 2-5 disks per hundred a year.
But the bad, wrong operating conditions raise this level much higher. Up to 10-30 per year.
. And you will not even understand with what it is connected specifically.

The first and most influential faults are vibrations of various kinds.
Anything can cause dangerous vibrations, a compressor in the next building, clapping the door two floors above.
The resonance of the rack itself.

The third is the temperature regime, closely related to pressure and humidity. The third one is the different power supply interference. Therefore, if the level of failures exceeds 5 disks per year per hundred, it is necessary to monitor the operating conditions by various measuring instruments and look for the cause.
Because even one random vibration can kill dozens of discs, but not immediately, but a few months later. While the first small scratches will break up to failure.
25.04.2017 12:26:00

If the heaps are flying,
This is one shelf. There is one heap and not one.
25.04.2017 12:36:00

If there are heaps, then something is wrong with the conditions.
What can be wrong with the conditions in the data center with air conditioning, guaranteed power, etc.? (No walkway with a slammed door there, of course, is also not observed)
The old disks there with the same speed for some reason did not fly. . .
25.04.2017 12:59:00
That's garbage, only with Dell and specifically with Seagate Cheetah 15K disks. 7, they use them for OEM. Only the problem is clear and understandable in what - without lead solder. Even on new disks, only with packaging, traces of the oxide around the bolts are already visible.
25.04.2017 13:00:00
Speech about one shelf.
There is one heap and not one.
I sometimes see different racks, and I think what kind of nerd it was developing.
The simplest ventilation rules are not observed.
Not to mention any protection from vibration and resonance.




old drive there is the same at this rate for some reason did not fly. . .
Well, they were quite different.
Order from the manufacturer discs up to 500 gig, without perpendicular recording, AF and other modern delights.
Once you can not provide the necessary conditions for new drives.
Inserting new disks into ancient racks, they are not calculated, are not always a good solution.
25.04.2017 13:13:00
Well, they were quite different.
We have disks running tens of thousands and for some reason " Perpendicular recording, AF and other modern delights " By no means prevent them from working. Marriage as was 10-20 years ago at the level of + -1%, so it is.
The author has one shelf - it's 12-24 discs (with rare exceptions). We can assume that the pieces 2-3-4 died (which fits perfectly into the banal bad luck or the notorious loader with the box) - but extrapolation already to the whole universe. B2039BDDDEFEF4FF68794712D8011B9A5
25.04.2017 13:26:00

So I'm saying that factory conditions of operation give failures at a couple of percent.
At a specific location, usually no more than 5%.

If there are more failures, search for the reason in your country or in the delivery conditions.
This is not always the case on the surface.
The reason may be the most unexpected, or when you do not think that because of it there may be an increased level of failures.
26.04.2017 16:00:00
quote:
:

Only here ST31500341AS cram in a modern computer without problems, and may already be a RAM of 2008 be some technical difficulties

Plus if The 11th series with buggy firmware SS at once To release.
26.04.2017 17:55:00
lololololo lololololo
ST31500341AS

Plus if The 11th series with buggy firmware SS, the right to emit. The problem is that this screw belongs to the Brinks family. And your " Bugfix firmware cc "
26.04.2017 19:34:00
Once discs of different years are considered, it was possible to write in the table for years - not everyone knows it by heart.
26.04.2017 19:41:00

quote:
as falling Reliability modern HDD
Reliability IDE drives decreases faster. . .
reading this forum and I realized that the golden days of the reliability of IDE drives have sunk into oblivion
10. 03. 2003
Now all new screws - D full
11. 03. 2003
Well, then

lololololo lololololo
B7A964F0E4A640779635064E836A 0AB9 quote:
Plus if The 11th series with buggy firmware SS, the right to emit
Who would want replaced on neglyuchnuyu
or six still somewhere so about five years ago. . . I do not remember
27.04.2017 9:03:00

Inserting new disks into ancient racks, which are not calculated on them, are not always a good solution.
And where did you get the fact that we have ancient stands?


The author of one shelf
see here psychic telepathy. . . However, in this case you are a little mistaken. . . There are several shelves.

It can be assumed that the pieces 2-3-4
4 died in less than six months from the last delivery of 8 terabyte disks. Somehow bust I think. . .
27.04.2017 9:29:00

And where did you get the fact that we have ancient stands?
And what's the difference.
Old or new, but made for the same requirements as the ancient racks
Pure practice, which shows that most developers spit on basic rules.

Find on the Internet you can already tell the ancient books on the general rules of ventilation electronics.
And see if your rack meets these requirements.
If not, additional ventilation for the Stoics must be provided, and not for the entire room.
I do not remember all the details. For a long time abandoned new developments, but only " Chinyu " Wretched burzhuyu junk.
There are gosts, requirements and tests for this correspondence.
In general terms, the distances between the heating blocks, in our case by disks, are not less than a centimeter.
Ventilation holes not less than 5 mm.
Well, the exact dimensions of the ventilation ducts are considered by formulas based on the heat transfer of the blocks.
For discs, the temperature gradient is also very important.
And this does not attract anyone's attention at all.

There are also requirements and rules for devices that are sensitive to vibrations.
The permissible vibration for disks is of the order of 0. 2 G. And this is very small.

And the correct fight against interference in the power supply in the multi-block rack, this is generally aerobatics for the designer.

All this is often spit, based only on commercial interests.
Roughly, if the rack was developed for terabyte disks, and you put a fresh 8 TB.
That and the requirements for operating conditions have been tightened at least 2 times.
27.04.2017 10:35:00
For the contrast, it would have been quite ancient to take gig at 60
27.04.2017 11:02:00

4 less than six months
Infant mortality, it's okay.
What kind of shelves?


There are several shelves.
Well this you wrote in the singular

27.04.2017 11:26:00

Find on the Internet you can already tell the ancient books on the general rules of ventilation electronics.
And with what reason should I look on the Internet for something that is not a fact that exists at all? If you are referring to something, then be kind enough to do this, and not shift the burden of proving your applications to the one to whom you are applying.

In general, the distances between the heating blocks, in our case by disks, are not less than a centimeter.
In which modern disk arrays, the distance between disks is more than a centimeter?
For example, HP StorageWorks D2700 and the like for use are not suitable, did I understand your statement correctly?
27.04.2017 11:46:00

HP StorageWorks D2700
The hoolet, like other brands, comes with a lot of whipping from the box. Therefore for decent clients it is necessary to drive iron pieces through the output control. . . Half of the beating is caught. . .
27.04.2017 11:50:00

Zdraste!
You have the problem you need. And do I have to prove something to you?
Zashibis.
I showed you the direction to move and what to look for.
And really work yourself
There are things that you just need to know from the course of general electronics and general principles of designing any equipment. And the evidence is not what they require.
Either you do not have this education, or you slept in lectures.
There is only a question of who took you to work, if you are not aware of such basic things.
27.04.2017 12:00:00

Take it easy on the corners.
The author does not have blackbase, but quite decent gamblers.
27.04.2017 12:03:00

And what HP does not mow? Yes, as much as necessary.
Hardy, which the author uses, is in the list checked with this rack and recommended by the manufacturer?
If not, then there's nobody to blame but the author.
27.04.2017 12:04:00

In general, the distance between the heating units, in this case, the disks are not less than a centimeter
vents are less than 5 mm
And that's all what the requirements for disk shelves or something?


The hard drives that the author uses are in the list checked with this rack and recommended by the manufacturer?
Are the discs compatible with the cabinet? , You sbrendili?

Statistics for those interested.
Three HP storage systems EVA, disks are no longer enterprise - 15K 2. 5 " SAS. Totally 666 disks. Within 4 years of operation, died 22.

27.04.2017 12:08:00

It is understood. You also do not know about the general rules of ventilation.
And here are the disks, when are these general rules for any electronics?

I give a hint, the rules are set out mainly in the documentation for calculating the design of various radiators for cooling the electronic components.
The disk does not differ in this respect from any other electronics.

27.04.2017 12:13:00

And where are the disks, when are these general rules for any electronics?
What does any electronics have if the discs are in a particular shelf with their fans? But they do not fit stacked on the floor of the cabinet.
27.04.2017 12:15:00

You are also not aware of the general rules of ventilation. Ostapa suffered from the loss of D13367732BDD489C894DD1D9A0EEB40C
.
27.04.2017 12:16:00

For 4 years of operation, died 22
A, not, 23rd on the way.
27.04.2017 12:22:00

Three HP storage systems, EVA disks, there are now no disks available - 15K 2. 5 " . Totally 666 disks. For 4 years of operation, died 22.
Well, this is absolutely normal operating conditions.
22/666/4 * 100 = 0, 8% per year.
And if
4/8 / (0, 5 * 2) * 100 = 50%
Either obvious problems in the operating environment,
or the left supply, for example, recovered and recovered software discs.
27.04.2017 12:28:00

22/666
Then / 688 (666 + 22).
27.04.2017 12:32:00

These are subtleties that you must correctly give.
Maybe you have 100 left in stock, they also need to be considered.

Sampling of less than 100 hard drives, and periods of less than a year, are usually not considered presentable.
This may be influenced by completely unrelated events.
The ceiling installer, for example, was in synthetic pants, but stood on the counter instead of a ladder.
28.04.2017 10:33:00

In the hoolet, as with other brands, quite a lot of shaving comes from the box.
Maybe it is. . . 2 screws so in general in one day perished. Well at least RAID was the 6th, which helped to avoid big problems (some of them certainly were, the array was rebuilt for a long time, which caused quite noticeable disruptions in the work due to the performance drop).
We will look further. . .


You need to. And do I have to prove something to you?
If you are saying something, then yes, the burden of proof lies with you. Otherwise, you can declare that neither be of the type " Earth is flat " , And to prove is your problem.

There are things that you just need to know from the course of general electronics and general design principles of any instrument.
Knowledge on something should be based, otherwise it's blind faith, and it has nothing to do with technology.
I asked you a very specific question: " What is the basis of your statement that the distance between disks in storage must be at least a centimeter? ? I asked you to give an example, correct from your point of view of modern storage systems.
You also hit the lyrics with the transition to the person with no specificity from you to achieve and failed.

The hard drives that the author uses are in the list checked with this rack and recommended by the manufacturer? Moreover, they were delivered together with the system.
.
28.04.2017 14:52:00
I can say this - in the HostingCommunity (a group of companies RuCenter, Hosting Center RBC (HostingCentre), GarantParkTelecom, Spaceweb, PeterHost P-01, Reggie after 2013) was vetoed for the purchase of server drives Sigate. Only WD or Hitachi. Regardless of which of the data centers the server will go. The reason is simple - the massive output of the server Sigeytes is out of order in 2011-12.
28.04.2017 15:56:00

If you are saying something, then yes, the burden of proof lies with you. Otherwise, you can declare that neither be of the type " Earth is flat " , And to prove is your problem.
ie type in the type search for such a query
https: // yandex. Com / search /? Text =% D0% 9F% D1% 80% D0% B0% D0 ...% D0% B2 & amp; Lr = 213
and read - above your capabilities?

Specifically about a centimeter I'm unlikely to find a document.
This can be said the general advice of all experienced designers.
You can not count accurately, due to a lack of all data, take at least a centimeter in the stock.
28.04.2017 16:03:00

Specifically about a centimeter I'm unlikely to find a document. - because with compulsory blowing a device 10cm long this is nonsense. Do not confuse with natural convection.
28.04.2017 16:31:00

To make sure that everywhere everything is correctly blown, you need to measure the temperature everywhere. Is there any flow of air there at all?
Therefore, he said that if there are problems, you need to measure everything and check. I have a lot of experience in EC technology.
There under each shelf, the rack had two powerful fans, both on top and bottom.
And God forbid the cover of the blocks to close, where the theses stand.
The convention is violated, FEZs are bent due to uneven cooling and no contacts start.
If you think that all these rules do not work for hard now. That is deeply mistaken.
Type found with what to compare, with the technique of fifty years ago.
Laws of Nature or who did not abolish.
In modern hardwoods, the head height adjustment is applied.
That basically was not 9 years ago
And it depends on temperature and on the slightest fluctuations of pressure and vibrations. Somewhat not so, immediately get scratches on the surface of the plates.
28.04.2017 16:46:00
That is deeply mistaken.
28.04.2017 17:29:00

https: // yandex. Com / search /? Text =% D0% 9F% D1% 80% D0% B0% D0 ...% D0% B2 & amp; Lr = 213
and read - above your capabilities?

So I thought. . . A typical " Google and look for me for proof that I've written you here " . . .

Specifically about a centimeter, I hardly find you a document.
And specifically about " Correct " On your parameters of storage, as far as I understand, too.
Everything is clear with you, you can not continue. . .

I have a lot of experience in EC technology.
TEAs bend due to uneven cooling and no contacts start.
29.04.2017 0:46:00

And specifically about " Correct " On your parameters of storage, as far as I understand, too.
And this is not my field of activity to choose the right rack.

You are stubbornly think that the discs you put a mess.
I told you about the fact that a crappy disk can not practically leave the factory.
And the failure rate should not exceed a few percent.
Unless you intentionally shove bad, untested discs into boxes.
It is tested so much that it is practically impossible.
After assembling yourself, " Creates " So to speak.
Applies a servo, creates zonal tables, factory defects sheets. Measures levels of acceptable errors and creates certification logs. If he's shitty after the assembly, he basically can not do it.
This leaves only the options:
- Substitution for recovered after failure. An unscrupulous supplier could put such disks to you.
This can be determined by the diagnostic utilities before the operation is started.
- Very poor delivery conditions, carried on ruts, not observing any climatic conditions.
It is easy to check the status of the various components of the structure and contacts.
- Poor operating conditions.
You do not want to check anything out of this. Of course it's easier to assume that the discs are rubbish and do not strain
And it's impossible to help you.
02.05.2017 9:06:00

The hoolet, like other brands, comes with a lot of whipping from the box.
Today, another one died (well at least 2 as soon as last time). Apparently, the whole party is defective. It hurts the process continues for a long time. . .



And this is not my area of ​​activity in order to choose the correct stand.
That's it! That is, in the matter under discussion, you do not even understand? Why then did my brain make so much of my invented " Standards " etc. ?
Like to prove to the people around what they are insignificant compared to you in the flames of a lot of things there. Why else in the technical with this climb?


I'm telling you that a crappy disc can not practically exit the factory.
Tell this to those who have actively used IBM DTLA, Fujitsu MPG and others so much " Reliable " Models, Mr. " Specialist " . .
02.05.2017 11:51:00

I tell you one thing, you tell me something else.
You need to find out for yourself, as a specialist, why you have so many failures.
And all the options I gave you.
The rest is just a lyric, as it happens, that producers rudely break elementary recommendations.
You want me to give you the results of running out thousands of different devices and chose the most reliable option for you.
And this is not possible

DLTA and MPG are the result of a series run-in.
If you were in the subject, you would know that any new feature is tested first in theory, then on experimental samples, on a test batch and on a serial run-in.
Constantly removing the identified problems from the test results.
Not all faults are detected immediately, therefore the term of successful tests for a new development is established already in 5 years.
Previously, it was senseless to say that the product was really reliable.
And you as a hurry, give me immediately and very reliably.
And here it is necessary, as they say, to trust but to check.
Your discs could be initially fucked up, and it surely would show tests.
And no serious expert will immediately deliver a new device for serious work. And will hold tests at least for a month, on non-responsive or test tasks.
At least one disc has been dropped during the test, it is necessary to extend the test period.

And this is what I wrote, most current specialists with paid diplomas do not even know.
The remnants of common sense, and teachers in institutes and heads of all levels, were completely lost.
02.05.2017 13:08:00

And no serious expert will immediately deliver a new device for serious work. And will test at least a month, on non-responsive or test problems.
Very, very interesting. Znachitstsa, comes to me bought for some sort of storage task, and I have to twirl it for an unreasonable task for months? And with the one, the responsible task for which she was buying, what should she do? Waiting?
Pages: 1 2
HDD, SSDForum → Discussion of the article: " Seagate Winchesters from Barracuda 7200. 11 to BarraCuda "

Discussion of the article: " Seagate Winchesters from Barracuda 7200. 11 to BarraCuda "

info@www.hdd-and-ssd.com