![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
2TB Reliability and Testing?
- Thread starter defuse
- Start date
drescherjm
[H]F Junkie
- Joined
- Nov 19, 2008
- Messages
- 14,941
At home and at work I use badblocks on just about every spinning disk I get.
This is available on any linux livecd. Allow it to do a 4 pass read /write test. That should take 20 to 30 hours for a 2TB drive. After the test if there are no badblocks reported at all look at the SMART raw data to see if any sectors were reallocated.
An example running of badblocks is
Where /dev/mynewdrive is the device you want to test.
If you are unfamiliar with linux its best to pull the SATA cables on all other drives in the system because this will destroy all data on the target disk without any warning. And even then verify you have the correct device.
hdparm -I /dev/sda
or
smartctl --all /dev/sda
should give you the serial # of what linux is using for /dev/sda
Remember that devices in linux will most likely be /dev/sda, /dev/sdb, /dev/sdc ...
If you are uncomfortable with linux a few full formats in winddows followed by looking at the raw SMART data will also work. I recommend using CrystalDiskInfo to look at the SMART.
This is available on any linux livecd. Allow it to do a 4 pass read /write test. That should take 20 to 30 hours for a 2TB drive. After the test if there are no badblocks reported at all look at the SMART raw data to see if any sectors were reallocated.
An example running of badblocks is
Code:
badblocks -wsv /dev/mynewdriveWhere /dev/mynewdrive is the device you want to test.
If you are unfamiliar with linux its best to pull the SATA cables on all other drives in the system because this will destroy all data on the target disk without any warning. And even then verify you have the correct device.
hdparm -I /dev/sda
or
smartctl --all /dev/sda
should give you the serial # of what linux is using for /dev/sda
Remember that devices in linux will most likely be /dev/sda, /dev/sdb, /dev/sdc ...
If you are uncomfortable with linux a few full formats in winddows followed by looking at the raw SMART data will also work. I recommend using CrystalDiskInfo to look at the SMART.
A lot of that has to do with shipping. If newegg boxed it up well and the shipping company did not play football with the box you should be fine..
When ordering HDDs from Newegg, it's best to limit your purchase to 1 or 2 HDDs, and nothing else. When you just order 1-2 HDDs, they usually do a pretty good job of tightly packing the box.
But, if you order HDDs along with something else, like a motheboard, PSU, or video card, then they're likely to pack them all together in a larger box. The larger box allows:
1. The packing material to settle, allowing the items inside to rattle around.
2. Items inside the box to move around such that they end up right up against a corner or side.
Both 1 and 2 greatly increase the odds that something fragile like a HDD, will arrive broken or damaged.
You can also order 8ish or more drives, in which case they'll likely use the carton the manufacturer used to ship them the drives. The manufacturer cartons do a good job of cradling the drives and keeping them from moving around inside the box.
But, if you order HDDs along with something else, like a motheboard, PSU, or video card, then they're likely to pack them all together in a larger box. The larger box allows:
1. The packing material to settle, allowing the items inside to rattle around.
2. Items inside the box to move around such that they end up right up against a corner or side.
Both 1 and 2 greatly increase the odds that something fragile like a HDD, will arrive broken or damaged.
You can also order 8ish or more drives, in which case they'll likely use the carton the manufacturer used to ship them the drives. The manufacturer cartons do a good job of cradling the drives and keeping them from moving around inside the box.
Thanks drescherjm!!
stewartjm, I live in Socal maybe I should just do will call.
http://promotions.newegg.com/NEPro/11-1312/index.html
stewartjm, I live in Socal maybe I should just do will call.
http://promotions.newegg.com/NEPro/11-1312/index.html
In the grand scheme of things, the manufacturers are shipping tens of millions of these drives a year, and maybe a couple of hundred people complain over the years on newegg. And of them, how many were installed improperly, used for a purpose they were not intended or right for (eg OMG my drives keep dropping out of my array these drives suck, and they bought drives not made to be raided due to error recover issues, tler, etc), didn't have proper cooling, vibration problems, power problems etc. That is a bit different than a 50% failure rate.
Trepidati0n
[H]F Junkie
- Joined
- Oct 26, 2004
- Messages
- 9,269
Also...people always bitch more than they praise. Happy people use their stuff...
NoxTek
The Geek Redneck
- Joined
- May 27, 2002
- Messages
- 9,300
....and when you calculate that into the well known fact that 98% of the people who write Newegg feedback are complete fucking morons, well... you do the math.
that wont help in the long run because the hard drive will still break and you will lose the data on it. you need multiple different types of storage to store redundant copies if your data is to survive. this is called backup
I have used Bart's Stuff Test to brutalize new hard drives in many-hour stress tests. but I've still had passing drives fail months later.
I've used a similar testing regime for years and it's served me well. I run badblocks followed by a zero fill for 2-3 passes to stress test a drive. Through this initial test process you will usually identify most faulty drives or drives which may be prone to infant mortality (failure within the first 6 months).
It's also true that most HDD failures are caused by improper handling or shock which may damage internal components. Usually this damage is the result of poor packaging (something Newegg has a bad reputation for with OEM drives).
blaze15301
Limp Gawd
- Joined
- May 16, 2012
- Messages
- 380
i use hd tune. then i use a number of programs to right zeros. and i basically nuke the hell out of a drive.
While it is true that people are much more motivated to leave a review if they are UNHAPPY with a product (5 times more likely in fact), sometimes it can be an indication.
For instance, a certain Seagate 2TB drive got more than 2500 bad reviews in a week on newegg. There was a real firmware issue. Would I look at these reviews and take them to heart? Yeah. I waited 3 months and the problem got fixed, then i bought 4 drives. all 4 are still running.
Just know what you are getting into and read not just what reviews are saying, but how they say it. sometimes its obvious how much the reviewer actually knows about a product, and how much you can trust their advice.
Prophes0r
For instance, a certain Seagate 2TB drive got more than 2500 bad reviews in a week on newegg. There was a real firmware issue. Would I look at these reviews and take them to heart? Yeah. I waited 3 months and the problem got fixed, then i bought 4 drives. all 4 are still running.
Just know what you are getting into and read not just what reviews are saying, but how they say it. sometimes its obvious how much the reviewer actually knows about a product, and how much you can trust their advice.
Prophes0r
In the past 12 years personally I've only observed about five instances of engineering or firmware defects or severe quality control problems with hard drives that resulted in a high number of "bad apples" appearing in the distribution channel. I consider 2006-2009 for Seagate to mostly be one prolonged occurrence.
2001 - IBM Deskstar GXP line (aka Deathstar)
2004 - Maxtor 250GB drives (high failures, Dell dropped Maxtor as OEM supplier, drives re-distributed to retail partners Compusa, etc at reduced pricepoint $100)
2005 - December 21, 2005 - Seagate acquires Maxtor
2006 - Seagate quality control for manufacturing and firmware microcode worsens with 7200.10 series(Seagate opens China plants during this time period)
2009 - Jan 19, 2009 - Seagate acknowledges firmware issues on 7200.11 series drives, releases 4 firmware updates in succession and manages to brick consumer drives in the process.
2010 - Samsung F3 & F4 - "Identify device" command results in bad block creation which could result in data loss.