Printable View
Anvil,
That's pretty unusual now... That problem drove me crazy even after I fixed the 'random' crashes. ASU will still pause -- lets say around 5000 files deleted -- for several seconds during which the whole system freezes even though the Mushkin isn't the boot drive. Since 3.3.2 the crashes seems to be gone, but there are other factors at work that might not manifest with the current paradigm of ASU, static data, and files generated.
Very unusual, it must be 30+ days since last time I had issues.
As I cant find anything in the Event log I'm assuming it was the F3, it is the most likely component and coupled with the 0 files I'm almost 100% confident in my conclusion :)
I will say that the Mushkin was extremely sensitive to that particular issue and with the 3.2.0FW I had considerably worse luck with that and Beta 9 together. I'm still running Beta8 at the moment, so maybe I'll try it again now that the 3.3.2. seems to have worked it's magic. There are still a few configurations I could try to coax the issues out, but I'm actually liking just having the system running like it should to the tune of 10.7GB a day.
If anyone out there knows anything about the new revised Plextor drives beyond what I see in the retail product listings, let me know.
UPDATE
Wow, I really should not have said any of that. I just logged into my system and the Mushkin has disconnected. ASU was set for 11GiB min free space, and it crashed with 11.01GiB free... It most likely happened in the past hour. Why did I have to go and open my big mouth? I have a screen shot I'll get posted up later.
UPDATE 2
I think the Mushkin just died.......
It didn't disconnect. It died, and no method will restore it to life at the moment.
https://www.box.net/shared/static/h7...yn5g4tteuf.jpg
I'm trying to figure out what happened and if there's anything I can do about it.
OK, so I've tried a couple of systems and the Mushkin can't get detected in any of them.
Here is a screen cap of the most recent update: 12 hours old from yesterday, day 50.
https://www.box.net/shared/static/8a...3slf1ent2v.jpg
Wow, that really is disturbing news.
Are there any LEDs?
There seems to be one red LED inside the chassis, one that stays solid. It's so hard to tell though. Its very difficult to see as well as you have to look for it through the space around the SATA connectors.
I did try it in a USB enclosure too, but to no avail. I'm going to leave it disconnected for a few hours and then try it again.
Annoyingly, I've had my first two dead SSDs in two days and neither one was what I expected. The Mushkin has just perished mysteriously, but not from exhaustion of PE cycles. The Agility died two days ago as well while running Windows XP on a Dell laptop. I'm having the Agility sent back to me so I can try to D-flash it before RMAing it.
I can't tell if the red LED is coming from inside the drive or not.
lol, so not one of the SF based SSD has managed to get past the finishing line. That failure sounds just like what occured to SymbiosVyse' drive. It’s all up to you now Anvil.
Except that mine had a WRD of 11, and only 11 reallocated sectors. Maybe there's something someone, somewhere, can do to save it.Quote:
Originally Posted by Ao1
M225->Vertex Turbo 64GB Update:
722.48 TiB (794.38 TB) total
1883.78 hrs (Torture), 2691 hrs (Power-On)
14238 Raw Wear
118.18 MB/s avg for the last 20.38 hours (on W7 x64)
MD5 OK
C4-Erase Failure Block Count (Realloc Sectors) from 13 to 14.
1=Bnk 6/Blk 2406
2=Bnk 3/Blk 3925
3=Bnk 0/Blk 1766
4=Bnk 0/Blk 829
5=Bnk 4/Blk 3191
6=Bnk 7/Blk 937
7=Bnk 7/Blk 1980
8=Bnk 7/Blk 442
9=Bnk 7/Blk 700
10=Bnk 2/Blk 1066
11=Bnk 7/Blck 85
12=Bnk 4/Blk 3192
13=Bnk 7/Blk 280
14=Bnk 3/Blk 2375
Alright, the Mushkin is really, really dead.
Any suggestions?
http://www.sadtrombone.com/Quote:
Originally Posted by Christopher
It's resting...
http://www.youtube.com/watch?v=4vuW6tQ0218
I have to say that I was not expecting that this is how the Chronos Deluxe would die. It looks as though CDI and Windows somehow really got it right and have been predicting failure for the past week. Basically, 01 (raw read error rate) would occasionally dip below threshold and trigger the move from CAUTION to BAD. Looks like there was something to it -- I just wish someone could adequately explain RRER in 2281s to me.
Is anyone else looking at their SF drives and getting apprehensive about the future?
I don't think anyone was expecting it to just fail, however it was well past the spec so we knew that it would fail, just not when.Quote:
Originally Posted by Christopher
Maybe we are too optimistic, maybe we need to give the drives a few breaks. (both powered on and powered off)
I'm not having second thoughts about my SF based drives, I'm pretty sure that none of my drives will get close to MWI exhaustion.
I expect I won't be using any drive that has expired for anything important, that goes for all makes.
I did however expect the drives to die gracefully, as in getting read only and not end up being undetectable.
Kingston SSDNow 40GB (X25-V)
466.50TB Host writes
Reallocated sectors : 05 12
Available Reserved Space : E8 99
MD5 OK
34.18MiB/s on avg (~25 hours)
--
Corsair Force 3 120GB
01 85/50 (Raw read error rate)
05 2 (Retired Block count)
B1 64 (Wear range delta)
E6 100 (Life curve status)
E7 10 (SSD Life left)
E9 344154 (Raw writes)
F1 458084 (Host writes)
MD5 OK
106.80MiB/s on avg (~25 hours)
power on hours : 1337
B1 is 1 down, from 65 to 64.
The only real indication that the drive was on the verge of dying was the periodic drops in RRER below threshold. That may just be a coincidence, but I still think that the Mushkin wasn't really all that far along, and certainly not past some point of no return. It didn't even have 380TB of raw writes on it yet. Perhaps you can only write so fast for so long to the drive before something (bad) happens. The M4 wasn't really all that much slower if you compare incompressible write speed, and with 25nm NAND as well.
I think this is more like a 'panic locked' Indilinx.
I guess I was just expecting a little more as toggle NAND drives are positioned as premium products with premium nand. If the drive had RAISE, maybe it would be a different story, or maybe it would be dead all the same. It just feels incredibly random, and that's not confidence inspiring. I wouldn't have a problem with using a SF drive for a boot drive, but I'm guessing that they are not well suited to surviving endurance testing. If the Force 3 has a similar fate to the Mushkin and the F40-A I think we need more of an explanation. I would consider this just rotten luck and not even related to the endurance testing had the F40-A died in exactly the same manner.
Todays update:
Kingston V+100
214.9147 TiB
847 hours
Avg speed 75.38 MiB/s
AD still 1.
P/E?
MD5 OK.
Reallocated sectors : 00
http://www.diskusjon.no/index.php?ap...tach_id=467288http://www.diskusjon.no/index.php?ap...tach_id=467287
Intel X25-M G1 80GB
33,4438 TiB Host writes
18810 hours
Reallocated sectors : 00
MWI=50 to 42
MD5 OK
52.25 MiB/s on avg
http://www.diskusjon.no/index.php?ap...tach_id=467289http://www.diskusjon.no/index.php?ap...tach_id=467286
Christopher,
When did RRER start dropping below threshold?
I won't change my plan for this drive but I'll be looking for signs of course, RAISE might make a difference but I do think I need to carry on in order to get some answers and then later on verify using a new drive. (we'll see)
In the past 6 or so days (I have email logs). I figured out that the RRER was dropping below the threshold which triggered CDIs emails and Windows' error log. It's hard to catch in the act as the value changes frequently, but I was watching when the value dropped below threshold and saw the drive's status change to BAD in CDI and was sent another few alerts. There has to be someone hanging out around here who could illuminate the Raw Read Error Rate 01 attribute as it pertains to SandForce 2200s.
I believe we have something useful out of the Muskin's death: RRER predicted the highly degraded state.
Possible root death causes as I see them:
- 1 * NAND depletion (Samsung 470 64GB)
- 1 * Data retention (Crucial M4 64GB)
- 1 * Uncorrectable errors (Mushkin Chronos Deluxe 60GB)
- 1 * Controller failure? (Corsair F40 40GB)
Out of which I believe Samsung 470 and Mushkin Chronos Deluxe were somehow predicted in a right way by SMART. Overall it's not bad, 50% chances to have a good death prediction. Probably hard drives are not too different from this point of view.
That may be true to an extent, but I don't think RRER is what it says with SF2200s. If you power the drive down for a short time, the RRER drops to zero when you power the drive back up. After a while of running it increases like a timer. Keeping better track of the attribute was difficult because it changes so often.Quote:
Originally Posted by sergiu
I bought a Larson Creek 311 20GB for cheap this evening. I'll check it out and may consider endurance testing it after putting it though its paces. I'm not crazy about testing SLC drives, but 20GB is probably small enough to be feasible. For 33% of MSRP I was buying even if I didn't need it.
An SLC drive should surely outlast what we have seen so far in this test.
20GB is not much but not that much of a difference vs the X25-E 32GB
It can do > 100MB/s sequentially but with this kind of workload it might be down to ~70MB/s.
--
Kingston SSDNow 40GB (X25-V)
467.97TB Host writes
Reallocated sectors : 05 12
Available Reserved Space : E8 99
MD5 OK
33.62MiB/s on avg (~38 hours)
--
Corsair Force 3 120GB
01 87/50 (Raw read error rate)
05 2 (Retired Block count)
B1 63 (Wear range delta)
E6 100 (Life curve status)
E7 10 (SSD Life left)
E9 347916 (Raw writes) ->340TiB
F1 463094 (Host writes) ->452TiB
MD5 OK
106.84MiB/s on avg (~38 hours)
power on hours : 1350
B1 is 1 down, from 64 to 63.
M225->Vertex Turbo 64GB Update:
731.19 TiB (803.96 TB) total
1904.10 hrs (Torture), 2714 hrs (Power-On)
14396 Raw Wear
118.39 MB/s avg for the last 18.85 hours (on W7 x64)
MD5 OK
C4-Erase Failure Block Count (Realloc Sectors) at 14.
1=Bnk 6/Blk 2406
2=Bnk 3/Blk 3925
3=Bnk 0/Blk 1766
4=Bnk 0/Blk 829
5=Bnk 4/Blk 3191
6=Bnk 7/Blk 937
7=Bnk 7/Blk 1980
8=Bnk 7/Blk 442
9=Bnk 7/Blk 700
10=Bnk 2/Blk 1066
11=Bnk 7/Blck 85
12=Bnk 4/Blk 3192
13=Bnk 7/Blk 280
14=Bnk 3/Blk 2375
That m225 Turbo is looking good. 800TB -- nothing to sneeze at.
Kingston SSDNow 40GB (X25-V)
469.44TB Host writes
Reallocated sectors : 05 12
Available Reserved Space : E8 99
MD5 OK
33.34MiB/s on avg (~51 hours)
--
Corsair Force 3 120GB
01 94/50 (Raw read error rate)
05 2 (Retired Block count)
B1 62 (Wear range delta)
E6 100 (Life curve status)
E7 10 (SSD Life left)
E9 351576 (Raw writes) ->343TiB
F1 467965 (Host writes) ->457TiB
MD5 OK
106.86MiB/s on avg (~51 hours)
power on hours : 1363
B1 is 1 down, from 63 to 62.
Todays update:
Kingston V+100
222.2637 TiB
875 hours
Avg speed 75.75 MiB/s
AD still 1.
P/E?
MD5 OK.
Reallocated sectors : 00
http://www.diskusjon.no/index.php?ap...tach_id=467422http://www.diskusjon.no/index.php?ap...tach_id=467421
Intel X25-M G1 80GB
37,6163 TiB Host writes
18838 hours
Reallocated sectors : 00
MWI=42 to 31
MD5 OK
49.23 MiB/s on avg
http://www.diskusjon.no/index.php?ap...tach_id=467423http://www.diskusjon.no/index.php?ap...tach_id=467420
Kingston SSDNow 40GB (X25-V)
470.71TB Host writes
Reallocated sectors : 05 12
Available Reserved Space : E8 99
MD5 OK
33.19MiB/s on avg (~62 hours)
--
Corsair Force 3 120GB
01 92/50 (Raw read error rate)
05 2 (Retired Block count)
B1 61 (Wear range delta)
E6 100 (Life curve status)
E7 10 (SSD Life left)
E9 354800 (Raw writes) ->346TiB
F1 472255 (Host writes) ->461TiB
MD5 OK
106.90MiB/s on avg (~62 hours)
power on hours : 1374
B1 is 1 down, from 62 to 61.
Todays update:
Kingston V+100
227.8350 TiB
897 hours
Avg speed 75.89 MiB/s
AD still 1.
P/E?
MD5 OK.
Reallocated sectors : 00
http://www.diskusjon.no/index.php?ap...tach_id=467613http://www.diskusjon.no/index.php?ap...tach_id=467612
Intel X25-M G1 80GB
41,2488 TiB
18860 hours
Reallocated sectors : 00
MWI=31 to 25
MD5 OK
53.77 MiB/s on avg
http://www.diskusjon.no/index.php?ap...tach_id=467614http://www.diskusjon.no/index.php?ap...tach_id=467611
Kingston SSDNow 40GB (X25-V)
472.19TB Host writes
Reallocated sectors : 05 12
Available Reserved Space : E8 99
MD5 OK
33.08MiB/s on avg (~76 hours)
--
Corsair Force 3 120GB
01 92/50 (Raw read error rate)
05 2 (Retired Block count)
B1 60 (Wear range delta)
E6 100 (Life curve status)
E7 10 (SSD Life left)
E9 358533 (Raw writes) ->350TiB
F1 477223 (Host writes) ->466TiB
MD5 OK
106.96MiB/s on avg (~76 hours)
power on hours : 1388
B1 is 1 down, from 61 to 60.
I installed a few updates last night -> system rebooted.
Cleaned the X25-V using the Toolbox and filled the F3 with 46% data and then deleted the file.
--
Kingston SSDNow 40GB (X25-V)
473.56TB Host writes
Reallocated sectors : 05 12
Available Reserved Space : E8 99
MD5 OK
36.76MiB/s on avg (~11 hours)
--
Corsair Force 3 120GB
01 94/50 (Raw read error rate)
05 2 (Retired Block count)
B1 59 (Wear range delta)
E6 100 (Life curve status)
E7 10 (SSD Life left)
E9 361629 (Raw writes) ->353TiB
F1 481343 (Host writes) ->470TiB
MD5 OK
106.56MiB/s on avg (~11 hours)
power on hours : 1399
B1 is 1 down, from 60 to 59.
I feel like I should be putting in some updates, but then I remember that the Chronos is dead... I blame Sandforce. They should make the drive with the understanding that people like me are going to write almost 11TB a day to it. :)
We miss the Chronos Deluxe and the m4!
Well, luckily there are new Indilinx drives coming soon! :DQuote:
Originally Posted by Christopher
And Samsung 830 :D
I'm somewhat dissatisfied by the death of the Mushkin (I mean, how much better would it been if the drive just exploded in a cacophonous shower of sparks?). I'm thinking about maybe getting another and trying again.
Is that the definition of insanity?
Todays update:
Kingston V+100
230.5225 TiB
911 hours
Avg speed 74.42 MiB/s
AD still 1.
168= 0 to 1 (SATA PHY Error Count)
P/E?
MD5 OK.
Reallocated sectors : 00
http://www.diskusjon.no/index.php?ap...tach_id=467798http://www.diskusjon.no/index.php?ap...tach_id=467797
Intel X25-M G1 80GB
45,2903 TiB
18881 Reallocated sectors : 00
MWI=25 to 18
MD5 OK
51.86 MiB/s on avg
http://www.diskusjon.no/index.php?ap...tach_id=467799http://www.diskusjon.no/index.php?ap...tach_id=467796
The Kingston has behaved strange today. First it dropped out like before. I restarted the rig and it was back again. I restored the log and started up again but I got an error message that Anvilpro.exe was broken.
http://www.diskusjon.no/index.php?ap...tach_id=467802
I deleted all files on the ssd and did a quick format before I copied it all back and restarted the test. Now it runs normal but ASU comes up with error messages.
Value 168 has changed from 0 to 1.
Anyone seen this kind of problem before?
That is half the fun :rofl:Quote:
Originally Posted by Christopher
What the insanity? :shocked::up:Quote:
Originally Posted by B.A.T
Yup :D
Kingston SSDNow 40GB (X25-V)
474.78TB Host writes
Reallocated sectors : 05 12
Available Reserved Space : E8 99
MD5 OK
34.78MiB/s on avg (~22 hours)
--
Corsair Force 3 120GB
01 92/50 (Raw read error rate)
05 2 (Retired Block count)
B1 59 (Wear range delta)
E6 100 (Life curve status)
E7 10 (SSD Life left)
E9 364685 (Raw writes) ->356TiB
F1 485413 (Host writes) ->474TiB
MD5 OK
106.73MiB/s on avg (~22 hours)
power on hours : 1410
No movement on B1.
--
The Force 3 is about to overtake the Intel. (the Intel's been writing for almost 6 months)
--
@B.A.T
I'd do a secure erase again, the messages looks to point to data corruption.
I would get Write Error 0 when the Mushkin would disconnect mid-loop, I would restart ASU and get WE0. I'd have to stop ASU, exit, then start again to get rid of it.
What is WE0?
WE0, Write Error 0
Okay. I've restarted and tried again but same problem. I'll try an SE and see if that helps.
Edit:
Now it runs just fine :)
isn't the Toshiba controller in that Kingston just a rebadged JMicron? Maybe I'm thinking of some other Kingston.
EDIT
Maybe I'm thinking of the new Kingston SATA III drives, the 200 series.
It's an Toshiba T6UG1XBG controller in the V+100
M225->Vertex Turbo 64GB Update:
759.98 TiB (835.61 TB) total
1971 hrs (Torture), 2784 hrs (Power-On)
14920 Raw Wear
118.36 MB/s avg for the last 70.46 hours (on W7 x64)
MD5 OK
C4-Erase Failure Block Count (Realloc Sectors) at 14.
1=Bnk 6/Blk 2406
2=Bnk 3/Blk 3925
3=Bnk 0/Blk 1766
4=Bnk 0/Blk 829
5=Bnk 4/Blk 3191
6=Bnk 7/Blk 937
7=Bnk 7/Blk 1980
8=Bnk 7/Blk 442
9=Bnk 7/Blk 700
10=Bnk 2/Blk 1066
11=Bnk 7/Blck 85
12=Bnk 4/Blk 3192
13=Bnk 7/Blk 280
14=Bnk 3/Blk 2375
Is that past M4 already?Quote:
Originally Posted by bluestang
Nope...last update before M4 (R.I.P) was 768 +/- TiB. So, once I hit 800 TiB, I'll call it past M4.
I love that M225-->Vertex Turbo.
It's a beautiful thing.
Yeah, not bad for a $1/GB refurb special from Crucial.
A data safety recap: how did each SSD fail (not detected at all, not writable, not readable at all, not readable fully..) and was there any degree of failure?
My drive, the Mushkin Chronos Deluxe 60 (SF2281 + 32nm Toshiba Toggle NAND) died suddenly and completely. Windows and CDI warned of the death of the drive a few days ahead of time, and I assume it had to do with attribute 01 RAW READ ERROR RATE (I refer to it as RRER). The SF1200 + 25nm Hynix NAND Corsair F40-A also died in a similar manner.Quote:
Originally Posted by alfaunits
The M4 was stopped for a few days, and upon restarting the system, the drive could not be found (this was believed to have been caused by super low data retention times, i.e. the M4 couldn't retain any data for the couple of days it was powered off).
I guess no one really knows for sure, but when you write as much and as fast as these drives are, it's hard for the NAND to retain any charge for long without power. Normally, a drive should be able to retain it's data for 1 year after exhaustion, but if you incur that exhaustion in a month or two, all bets are off.
Bluestang,
I bought a Vertex Turbo 128GB for 1$/GB, and they're still around (and not marked as refurbs). I'm thinking about buying another one while they're still around, as they're good for just about everything (though to be fair, your M225 Turbo is a Hall of Fame candidate for sure.)
Today was "RMA a lot of broken stuff" day.
Some companies are a lot easier to work with than others, and any company that does advanced RMAs in my book is excellent. I don't normally buy stuff from one vendor or another based on customer support, but I might have to reconsider.
Dealing with Asus is particularly irritating, and OCZ can be infuriating -- takes a long, long time to communicate. Fortunately, I've never had to RMA much stuff.
Mushkin, is really easy to deal with as you can just call them up for anything. You can buy stuff, get technical help or set up advanced RMAs in one short phone call (to an actual breathing person, even! Unlike some other companies I could name).
I bought a Chronos Deluxe 120 today from Mushkin, and I'm considering endurance testing it. It won't be here for a while, though. I'm also looking at some other 60GB SF2281s to see if it dies the same way the Deluxe 60 did.
I'm working on creating a summary in the first post, it will take some more time as the thread has grown a lot :)Quote:
Originally Posted by alfaunits
(I'm about half way there on collecting details, as soon as the details are collected I'll rewrite the first post completely)
@Christopher
Don't you know what day it is? :) -> SnB-E
Kingston SSDNow 40GB (X25-V)
477.49TB Host writes
Reallocated sectors : 05 12
Available Reserved Space : E8 99
MD5 OK
33.60MiB/s on avg (~46 hours)
--
Corsair Force 3 120GB
01 92/50 (Raw read error rate)
05 2 (Retired Block count)
B1 57 (Wear range delta)
E6 100 (Life curve status)
E7 10 (SSD Life left)
E9 371497 (Raw writes) ->363TiB
F1 494479 (Host writes) ->483TiB
MD5 OK
106.61MiB/s on avg (~46 hours)
power on hours : 1434
B1 has dropped to 57.
Todays update:
Kingston V+100
233.0828 TiB
935 hours
Avg speed 76.98 MiB/s
AD still 1.
168= 0 to 1 (SATA PHY Error Count)
P/E?
MD5 OK.
Reallocated sectors : 00
http://www.diskusjon.no/index.php?ap...tach_id=467961http://www.diskusjon.no/index.php?ap...tach_id=467960
Intel X25-M G1 80GB
46,9237 TiB
18890 hours
Reallocated sectors : 00
MWI=18 to 15
MD5 OK
48.26 MiB/s on avg
http://www.diskusjon.no/index.php?ap...tach_id=467962http://www.diskusjon.no/index.php?ap...tach_id=467959
The test has been down for some hours today due to a cpu upgrade but now it's running again :up:
SB-E Day?
More like "no Cherryville for you!!!" day.
I'm pretty sure that's what it's called on my calendar anyway.
B.A.T.
CPU upgrade? to what?
----------------------------------------------------------
Dear Intel,
If you want to just send me a new X79 setup, complete with DX79SI mobo and Extreme series processor, I'd be happy to use it to endurance test with. Failing that, just send me a couple of Cherryvilles and we'll call it even. I'll totally shill your gear! PM me!
If not, I'm going to rip you incessantly for still including PCI slots on you motherboards. I mean, it was your idea to ditch them, and now your "Extreme" mainboards have 1,2, and 3 PCI slots.
Really, it's up to you. I mean, everyone and their mom, dog, and obese hamster got review samples. Share the love.
Christopher
-----------------------------------------------------------
FX-8150. Got it in the mail same week my son came and finally I had the time and energy to do it :)
@B.A.T
So an FX-8150 for the Endurance test?
@Christopher
Did you notice the sale at Link :)
SnB-E :)
I've ordered the MB but have yet to decide on the CPU
Anvil,
I've been toying with the idea of picking up a Vertex LE. They're(well, the 100 anyway) actually a lot cheaper some where else... Actually, I got the Vertex Turbo 120 instead just because I liked the M225 "Turbo" so much.
You think they've got lifetime throttling?
I'm pretty sure they have Lifetime Throttling, still they are about the cheapest 34nm NAND you can get so they should be OK for normal usage.
It was for my main rig and at the moment I got the endurance test running from it so, yeah :)
I also updated my folding rig and I'm thinking of moving both ssd to that one instead.
And of course I ordrered another M4 64GB :rolleyes: It's demo/returned but for 680,- NOK I couldn't resist
I haven't decided what to do with it but maybe data retention or a rerun of the endurance test.....
I'm still thinking about running the M4 64GB I have. It's only used 3PE cycles, and I was thinking about how I would run it. Maybe overprovisioned? Maybe I could launch my testing rig into space for some cosmic ray testing...
So many options!
I haven't decided what I'm going to do yet, but I figured there should be at least one M4 in the test at all times.
Does NOK stand for Norwegian Okra? :D
You don't use slimy vegetables for currency do you?? :confused:
After looking in to how much some of these drives are in Norway, I'm glad I live in the land of cheap SSDs.
Cheap SSDs are the greatest thing ever.
Cosmic ray testing sounds like a lot of fun, you can take care of that aspect :rofl:
Our currency is Kroner and NOK stands for NOrwegian Kroner. 1USD=5,7 NOK
The M4 64GB cost 769 NOK = 134 USD (that is including VAT of 25%) I think newegg have the same one for 115USD.
I know what NOK stands for :)
I forget about the VAT every time.
That explains it.
The M4 has been getting more and more expensive here for some reason. I bought my M4 64 for $89 with shipping a few months back. Now the cheapest I've seen it it recently is $109 + shipping. The 128 could be found for as little as $165 for a time, now it's as expensive as the Mushkin Chronos Deluxe 120. And the Intel 510 is just stupid expensive now, especially compared to much cheaper drives. For the price of the 120 510 you can almost get an Agility 3 240. G1 and G2 Intel drives were going inexpensively on Ebay a few weeks ago, now they're all absurdly priced. I got a X25 E 32GB for $88 (I think that's somewhere around 500 NOK). Now when you can find them, they're four times that much. Pricing might be volatile for new drives, but used drives fluctuate constantly. I've bought 4 used Intel drives, and not one of them had more than 300GB of host writes on it. I've been looking at getting some more, but all the value is gone.
Buying a used SSD from anyone but us here in the endurance thread is a very low risk. They are hardly written to and with no moving parts to wear down the only problem are issues from production and you got warranties for that :)
Here are some ex from Norway
M4 64GB
http://www.prisjakt.no/bilder/graph_price.php?p=866383
M4 128GB
http://www.prisjakt.no/bilder/graph_price.php?p=873561
Intel 510 120GB
http://www.prisjakt.no/bilder/graph_price.php?p=819442
Intel 320 120GB
http://www.prisjakt.no/bilder/graph_price.php?p=844282
Vertex 3 MI 120GB
http://www.prisjakt.no/bilder/graph_price.php?p=867284
Although I'd want to get a guarantee that it has at least X% of its manufacturer specified erase cycles left, and if I got it and SMART indicated otherwise, I would be returning it.Quote:
Originally Posted by B.A.T
Low risk doesn't mean stupid :) A screenshot of CDI is never wrong when buying used.
I've had good luck buying used INTEL drives. For some reason, none of the 4 drive purchased a different times from different people had more than a few hundred hours, host writes, or power cycles. I'd be wary of buying used drives of other makes (but that's just my opinion). I can't say why my cross section of auction purchases have been almost new, but if a guy's splitting up his 8 drive array its a safe bet they haven't been thrashed too much (sometimes? Maybe stay away from Computurd's). If I can return the drive it makes me feel better about buying it.
I wouldn't want to buy a SF or Indilinx drive from a stranger, especially out of warranty. Especially OCZ Indilinxes as you can dflash them, misrepresenting actual writes and other factors when you get show some SMART data. Intel drives are just a safer bet when buying used. If you're buying second hand from a friend or someone you know, it's different.
I even opened up the drives when I got them to make sure they weren't some kinda Communist Chinese forgeries. Not sure how you make a fake X25-M, but I'm not taking any chances. Damn Communists and their counterfeit products!
I had a good run on Ebay, but mainly I just see a lot of absurdly high prices now. Occasionally, you'll see someone with an SSDlife/CDI shot on the auction site, but that's pretty rare.
I think what I'm trying to say is, if you're going to buy a used SSD from a stranger, drink as much Tequila as you can first. It will make things easier.
Sounds like a good advice :D
Kingston SSDNow 40GB (X25-V)
478.91TB Host writes
Reallocated sectors : 05 12
Available Reserved Space : E8 99
MD5 OK
33.37MiB/s on avg (~58 hours)
--
Corsair Force 3 120GB
01 95/50 (Raw read error rate)
05 2 (Retired Block count)
B1 56 (Wear range delta)
E6 100 (Life curve status)
E7 10 (SSD Life left)
E9 375057 (Raw writes) ->366TiB
F1 499220 (Host writes) ->488TiB
MD5 OK
106.60MiB/s on avg (~58 hours)
power on hours : 1446
B1 has dropped from 57 to 56.
It's getting closer to the first retention tests.
Oops, never mind me posting this :wasntme:
bluestang: Where is my daily dose of Vertex Turbo update?...:DQuote:
Originally Posted by bluestang
M225->Vertex Turbo 64GB Update:
768.91 TiB (845.42 TB) total
1992 hrs (Torture), 2810 hrs (Power-On)
15082 Raw Wear
118.42 MB/s avg for the last 20.87 hours (on W7 x64)
MD5 OK
C4-Erase Failure Block Count (Realloc Sectors) 14 to 15.
1=Bnk 6/Blk 2406
2=Bnk 3/Blk 3925
3=Bnk 0/Blk 1766
4=Bnk 0/Blk 829
5=Bnk 4/Blk 3191
6=Bnk 7/Blk 937
7=Bnk 7/Blk 1980
8=Bnk 7/Blk 442
9=Bnk 7/Blk 700
10=Bnk 2/Blk 1066
11=Bnk 7/Blck 85
12=Bnk 4/Blk 3192
13=Bnk 7/Blk 280
14=Bnk 3/Blk 2375
15=Bnk 7/Blk 768
Bank 7 has 7 bad Blocks now.
That's past the last known M4 state. Awesome!Quote:
Originally Posted by bluestang
Mushkin Chronos Deluxe 60 update, Day 56
FW 3.3.2
The Mushkin is still dead.
Still very sad about it too.
Mushkin, that was the most famous 60GB SF2281 in the world!!
Hi guys, long time no see.
It's a great thread you have here. I remember i wanted to test something like this back in '09, but the cost of SSDs then put it well outside what i could do.
I'm using this as a reference to SSDs being able to handle their rated 3000-5000 cycles in a report i'm writing. I've started a computer engineer degree in embedded systems now.
I tried donating one of my old Mtrons to BAT for endurance testing, but it was dead. Gonna try to RMA it since it's 3 years old but has a 5 year server warranty. The second Mtron of the pair still lives and does it's daily duty as boot drive in my little sisters laptop.
I've been thinking of starting another SSD project looking at sustained random write (4KB QD32) drop-off rate as a function of time and spare area (possibly also 10,25,50,75,90,100 % duty-cycles to allow some GC and compare effectiveness of it) for different SSDs. Have you guys done that while i've been away? If not, would you be interrested?
Nice to see you gullars!
if i had time and gear to put towards the testing, i would :)
I will definitely read and learn along with you guys though !
GullLars,
I remember your comments on AnandTech's X25-V RAID 0 review... they were really illuminating at the time. It's why I got a pair of X25-Vs in the first place... at least, I'm really sure that's where I know GullLars from. Maybe I imagined that.
Welcome back to the fold!
I don't know how up to date you are on this thread, but some of the SandForce drives had lifetime throttling, limiting their writes over the course of a year for longevity. Had the SandForce 2281 drive I ran (which died unexpectedly a few days ago) had Lifetime Throttling, I was planning on doing full span 4K writes as it would have surely come in under the throttling speed limit (maybe? never got that far). But that kind of testing really falls under the domain of enterprise-style testing. Though as it's turning out, just writing as fast as the drive can for weeks on end has some side effects you wouldn't experience if the writes were spread out over the span of a normal Consumer-based usage life.
That would be very interesting to see. If you are not already familiar with it, you may want to review the SNIA documents on SSD test procedures, since they should apply directly to the test you have described:Quote:
Originally Posted by GullLars
http://www.snia.org/tech_activities/..._standards/pts
Todays update:
Kingston V+100
238.8913 TiB
957 hours
Avg speed 76.06 MiB/s
AD still 1.
168= 0 to 1 (SATA PHY Error Count)
P/E?
MD5 OK.
Reallocated sectors : 00
http://www.diskusjon.no/index.php?ap...tach_id=468027http://www.diskusjon.no/index.php?ap...tach_id=468026
Intel X25-M G1 80GB
50,7363 TiB
18913 hours
Reallocated sectors : 00
MWI=15 to 07
MD5 =6 mismatches but OK after an SE.
49.16 MiB/s on avg
http://www.diskusjon.no/index.php?ap...tach_id=468028http://www.diskusjon.no/index.php?ap...tach_id=468025
Hi GullLars!Quote:
Originally Posted by GullLars
Ao1's been doing some initial testing at QD1 (full span, spare area), there is a separate thread for that. Link
It is indeed interesting!
I'm sure there is interest for more on this, a lot more.
--
Kingston SSDNow 40GB (X25-V)
480.27TB Host writes
Reallocated sectors : 05 12
Available Reserved Space : E8 99
MD5 OK
33.23MiB/s on avg (~70 hours)
--
Corsair Force 3 120GB
01 84/50 (Raw read error rate)
05 2 (Retired Block count)
B1 55 (Wear range delta)
E6 100 (Life curve status)
E7 10 (SSD Life left)
E9 378492 (Raw writes) ->370TiB
F1 503792 (Host writes) ->492TiB
MD5 OK
106.59MiB/s on avg (~70 hours)
power on hours : 1459
B1 has dropped from 56 to 55.
Its really a shame that SSD parts aren't user interchangeable... You can't switch out processors or NAND. How great would it be if they were socketed? I could drop in a SF controller in place of a Marvel or Indy, then swap some 51nm for the 34nm IMFT. Basically like my sound card is made to swap OPAMPS.
Kingston SSDNow 40GB (X25-V)
482.24TB Host writes
Reallocated sectors : 05 12
Available Reserved Space : E8 99
MD5 OK
33.08MiB/s on avg (~88 hours)
--
Corsair Force 3 120GB
01 82/50 (Raw read error rate)
05 2 (Retired Block count)
B1 54 (Wear range delta)
E6 100 (Life curve status)
E7 10 (SSD Life left)
E9 383450 (Raw writes) ->374TiB
F1 510394 (Host writes) ->498TiB
MD5 OK
106.58MiB/s on avg (~88 hours)
power on hours : 1476
B1 has dropped from 55 to 54.
Looks like a data retention test for the F3 is due tonight :)
M225->Vertex Turbo 64GB Update:
779.39 TiB (856.91 TB) total
2016 hrs (Torture), 2835 hrs (Power-On)
15272 Raw Wear
118.63 MB/s avg for the last 25.51 hours (on W7 x64)
MD5 OK
C4-Erase Failure Block Count (Realloc Sectors) at 15.
1=Bnk 6/Blk 2406
2=Bnk 3/Blk 3925
3=Bnk 0/Blk 1766
4=Bnk 0/Blk 829
5=Bnk 4/Blk 3191
6=Bnk 7/Blk 937
7=Bnk 7/Blk 1980
8=Bnk 7/Blk 442
9=Bnk 7/Blk 700
10=Bnk 2/Blk 1066
11=Bnk 7/Blck 85
12=Bnk 4/Blk 3192
13=Bnk 7/Blk 280
14=Bnk 3/Blk 2375
15=Bnk 7/Blk 768
you can't really have interchangeable controllers if they all have a different number of channels.
I get that. I was just staring at two dead drives and wishing I could assemble one working drive out of out of it.Quote:
Originally Posted by Boogerlad
Though there is one company in Austria, or maybe Australia, that had a pseudo-modular PCIe SSD.
Todays update:
Kingston V+100
245.2253 TiB
982 hours
Avg speed 76.17 MiB/s
AD still 1.
168= 0 to 1 (SATA PHY Error Count)
P/E?
MD5 OK.
Reallocated sectors : 00
http://www.diskusjon.no/index.php?ap...tach_id=468170http://www.diskusjon.no/index.php?ap...tach_id=468169
Intel X25-M G1 80GB
54,7732 TiB
18938 hours
Reallocated sectors : 00
MWI=07 to 01 to 115 to 113
MD5 =OK
52.17 MiB/s on avg
http://www.diskusjon.no/index.php?ap...tach_id=468171http://www.diskusjon.no/index.php?ap...tach_id=468168
https://www.box.net/shared/static/mx...zcnxkoeq0z.jpg
The 311 has nice heavy feel to it, which I wasn't expecting, and it's also a little faster too. I'm still on the fence about endurance testing it though. I was running some numbers and I think it would just take too damn long. I think it would take somewhere near 2 years, maybe not even then. Who knows? Surely it would last a little longer than the Mushkin, and that's a start.
It has more of a stainless steel top, but the same stupid 2mm spacer on it.
https://www.box.net/shared/static/s9...b0vjf65vj6.jpg
--------------------------------------------------------------------------
I've been digging in a little deeper as to why the Mushkin went T.U.
Raw Read Error Rate is ECC + uRAISE errors.
Retired block count is not actual blocks, but 100 * (retired blocks / minimum acceptable free space in blocks). Not sure how the value of 11 stacks up, but I believe that each retired block may be much more significant than I initially believed -- but I still don't think 11 is high.
Both drives have been disconnected for 11 hours, I had to do a quick check and they are both alive.
(nothing read or written except for SMART attributes)
They will stay disconnected for another 10-12 hours.
Attachments are still not working ?!
You could run Anvils Endurance app for say 2 days to 2 weeks in order to get some generalized data.Quote:
Originally Posted by Christopher
That info could be interesting for comparison to other Intels used here in their early usage.
My opinion that all info gathered could be useful.
The first Vertex tested gave us useful info even though throttling eliminated it from further testing.
Just a thought.
M225->Vertex Turbo 64GB Update:
788.68 TiB (867.16 TB) total
2038 hrs (Torture), 2858 hrs (Power-On)
15441 Raw Wear
118.32 MB/s avg for the last 22.84 hours (on W7 x64)
MD5 OK
C4-Erase Failure Block Count (Realloc Sectors) from 15 to 16.
1=Bnk 6/Blk 2406
2=Bnk 3/Blk 3925
3=Bnk 0/Blk 1766
4=Bnk 0/Blk 829
5=Bnk 4/Blk 3191
6=Bnk 7/Blk 937
7=Bnk 7/Blk 1980
8=Bnk 7/Blk 442
9=Bnk 7/Blk 700
10=Bnk 2/Blk 1066
11=Bnk 7/Blck 85
12=Bnk 4/Blk 3192
13=Bnk 7/Blk 280
14=Bnk 3/Blk 2375
15=Bnk 7/Blk 768
16=Bnk 7/Blk 765
Bank 7 now has 8 bad Blocks.
I’d agree. Based on observations to date expiry of the MWI can be projected with precision, so you only really need to know the extent of writes it takes to diminish the MWI by a couple of percenatge points to be able to establish how long the drive will last.Quote:
Originally Posted by Hopalong X
Sadly it looks like if you don’t want to risk losing your data you need to replace the SSD when the MWI expires. No doubt the MWI is conservative, in context to the chances of losing data once it has expired, but testing that would be extremely time consuming.
Ao1,
I think once MWI expires you have a choice to make. How fast did you get to 0 MWI? If the answer is 1 week, then you're going to have problems sooner rather than later. If it takes you 19 months to get to MWI 0, you're probably still golden. Somehow, I'm only writing about .3 to .7GB a day to my main system drive in my primary desktop over the course of 8 to 12 hours a day that this particular system is on. But really, if you can write under 100GB a day to your drive, I think it should last quite some time. Should you find yourself writing 10+TB a day to a drive, it's not going to last long in any event. Had I spread out the 486TB host writes of the Mushkin over the span of 12 -24 months, it wouldn't really be a problem and would most likely make it to the end of warranty. Large numbers of host writes probably aren't a problem unless they pile up quickly.
Some of the drives, like the 40GB Intel units, just can't write fast enough to commit suicide, while other seem to be heading inexorably towards disaster. Maybe there is some sound logic to OCZ's insistence that SF drives have LTT.
My first planned data retention test is finally over :)
First test lasted for 11 hours, no issues, the next test lasted fore 24 hours, both drives looks to be fine, no issues so far :)
I'll think about how and when the next test will be performed but I expect in about a week or so I'll leave one drive at a time off for 48 hours and then the next week for 72.
We can't really expect to find out if the drive will last disconnected for months or years so I think I'll settle for finding when the drive can't take the weekend off.
No, it is the number of erases that are primarily responsible for degrading the flash and reducing data retention time, not the frequency of erases. While there might be a slight additional degradation from a large number of erases in a short time, the effect is most likely small compared to degradation from number of erases.Quote:
Originally Posted by Christopher
Unfortunately, data retention will be reduced significantly based primarily upon the number of erases performed on the flash. Assuming anything else is just wishful thinking.
I was speaking in terms of bit errors, the kind that I believe killed the Mushkin. I don't believe it was a data retention issue since the drive had not been turned off in two weeks by the time it died. Flash degradation plays a part, but I do not think that the flash wore out all at once. The Raw Read Error Rate has everything to do with how long and how fast you write. The more rest time in between writes, the fewer UCC+URAISE errors.
primarily would be true, but there is a big impact of idle time on the condition of nand. I believe that AO1 posted a whitepaper that shows that nand endurance can be increased exponentially by having idle time. I will try to find it, but i believe if i recall correctly that some went from 5k p/e ratio to 30k p/e ratio.Quote:
Originally Posted by johnw