#367158 - 05/07/2016 20:59
Drobo HDDs: high failure rate?
|
carpal tunnel
Registered: 18/06/2001
Posts: 2504
Loc: Roma, Italy
|
Guys,
I've had this Drobo 5N for more than two years now, and overall I am happy with it.
But, I noticed a really high drive failure rate, at least compared with my PCs HDD failure rate over the years. I am starting to suspect it may be the Drobo unit to kill HDDs sooner than my PCs - maybe due to temperature, or other parameters.
Do you guys think this may be the case? Has anyone had similar experience with Drobo units?
_________________________
= Taym = MK2a #040103216 * 100Gb *All/Colors* Radio * 3.0a11 * Hijack = taympeg
|
Top
|
|
|
|
#367159 - 05/07/2016 21:56
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 20/12/1999
Posts: 31591
Loc: Seattle, WA
|
If it's got a fan on the drives, as long as the fan is working, I would think that would be good. Is the fan working? (I've had fans die on me before, which tends to kill a bank of drives PDQ.) What kind of drives? https://www.backblaze.com/blog/hard-drive-reliability-stats-q1-2016/
|
Top
|
|
|
|
#367162 - 06/07/2016 07:10
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
pooh-bah
Registered: 12/01/2002
Posts: 2009
Loc: Brisbane, Australia
|
What temperature are they reporting when running?
How do they fail? The usual clunking kind of failure?
_________________________
Christian #40104192 120Gb (no longer in my E36 M3, won't fit the E46 M3)
|
Top
|
|
|
|
#367165 - 06/07/2016 14:33
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 08/03/2000
Posts: 12332
Loc: Sterling, VA
|
You also didn't mention what that failure rate was. Has it been two drives in two years? 5?
I bought my Synology two years ago this week, and I've never had a drive fail on me (*knocking on wood*). It only has four drives and for a long while it only had three, but I do have a couple video cameras recording to it at all times, and I've set it up for headless Crashplan (a real PITA), so it gets some work.
_________________________
Matt
|
Top
|
|
|
|
#367166 - 06/07/2016 15:48
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 18/06/2001
Posts: 2504
Loc: Roma, Italy
|
Guys, you're right. Here's more details:
1. Fan is there, it is working, and it is as noisy as always. I've been planning to replace it for a while, incidentally, with a quieter Noctua.
2. WD 1TB, 2TB and 3TB Green drives. Various subversions of them
3. I am afraid Drobo dashboard does not report drive temp.
4. Not sure HOW they fail. RED led next to the failed drive. I've just been replacing the drives without further investigation.
5. Failure rate: in 3 years, I replaced possibly 4/5 drives.
6. Admittedly, these are all drives I've previously used on my server at home, for at least 1 year. As I upgrade the server storage capacity, replaced drives end up in the Drobo, so none of its drive is new. One could argue they would've failed in the server at some point. However, other than a recent 8TB drive failed on me (see another thread on this board about drive temperatures), I have not had a drive fail in my server for 10 years at least, or more.
Matt, yes, I've been using Synology NAs units at work for various projects and, at least those I've been working with, they seem much greater quality than the Drobo. Mostly just an impression I get, admittedly, maybe because I like Synology OS so much better than the extremely basic Drobo's, but still.
_________________________
= Taym = MK2a #040103216 * 100Gb *All/Colors* Radio * 3.0a11 * Hijack = taympeg
|
Top
|
|
|
|
#367167 - 06/07/2016 16:09
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 08/03/2000
Posts: 12332
Loc: Sterling, VA
|
Wow, that does seem like way too high a failure rate. I don't think I've had 5 drives fail on me since I started using computers, let alone in the last three years! Have you contacted Drobo about this?
I can't really speak to Drobo, as I've only set up one device, but everything you said about them seems accurate to me. The construction doesn't inspire much confidence, and the software seemed pretty bad. I liken it to my experience with Carbonite vs Crashplan. Carbonite's software, IMO, is garbage and hardly gives you any options. Crashplan's software lets you do a lot more, and while it might not be as good as I want it to be, it's way better.
But none of this is helpful, so I apologize.
I wonder what Drobo is using to determine a bad drive. Is it reading the SMART data? Is the drive fully failed? The next time a drive fails (and that seems to be a pretty good certainty), you might try removing it and then hooking it up to use as a junk drive for data you don't care about losing and see how long it lasts. You could use it as a tertiary backup drive that you don't mind sacrificing. If it's usable, I'd just be curious to see how long it lasts.
_________________________
Matt
|
Top
|
|
|
|
#367168 - 06/07/2016 16:12
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 29/08/2000
Posts: 14489
Loc: Canada
|
I don't know about the Drobo, but a problem with nearly all of the small "home sized" NAS units that I have seen, is they simply don't keep the drives cool enough. Fan or no fan, the airflow design is nearly non-existent.
So I would bet on overheating, leading to media errors (aka. "bad sectors").
|
Top
|
|
|
|
#367169 - 06/07/2016 16:59
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 20/12/1999
Posts: 31591
Loc: Seattle, WA
|
2. WD 1TB, 2TB and 3TB Green drives. Various subversions of them Hm. I had two WD 2TB drives fail in quick succession in my Synology NAS recently. I wonder if there's something to this. I am afraid Drobo dashboard does not report drive temp. Feel 'em with your finger. Are they like OMFG hot, or just nicely warm? 4. Not sure HOW they fail. RED led next to the failed drive. I've just been replacing the drives without further investigation. Might be worth connecting the failed drives into a PC and using a S.M.A.R.T utility to read stats. Maybe there's nothing physically wrong with the drives and the Drobo firmware has some kind of a partition-mangling bug.
|
Top
|
|
|
|
#367170 - 06/07/2016 17:52
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 18/06/2001
Posts: 2504
Loc: Roma, Italy
|
Thank you all, guys, you're amazing as always.
So, I very much suspect temp is an issue here. Every failed drive I pulled out of the box was nearly "OMFG hot". I say "nearly" just because occasionally I've touched hotter drives, but yes, these are on the very hot end of the spectrum of my personal experience. So, yes, great point, cooling is not ideal for sure. Also, I started this thread because one drive just failed on me, yesterday, and interestingly it happened when temperature in Rome rised significantly and, in spite of AC at home, temp in the room where the drobo also increased.
Also, as you are correctly noticing, I've also been wondering if Drobo is being too conservative here and reports a drive as Failed when a PC or other NAS models would probably not. Only once I've tried to connect a drive to a PC (I should've done that more often), and interestingly it was recognized and I could format it. But, it became unresponsive the day after, as I tried to stress it a bit by running a backup script TO it. So, I decided it was dead and did not spend more time analyzing. However, latest failed drive is till at home on my desk, I'll give it a try and see if it is in fact still alive.
I suspect there's a combination of factors here, maybe:
1. Poor cooling 2. Possibly, conservative approach at marking a drive as "Failed" 3. Old drivers being used in my particular case.
_________________________
= Taym = MK2a #040103216 * 100Gb *All/Colors* Radio * 3.0a11 * Hijack = taympeg
|
Top
|
|
|
|
#367171 - 06/07/2016 17:55
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 20/12/1999
Posts: 31591
Loc: Seattle, WA
|
Did a fan go bad in your Drobo?
|
Top
|
|
|
|
#367172 - 06/07/2016 20:06
Re: Drobo HDDs: high failure rate?
[Re: mlord]
|
carpal tunnel
Registered: 08/03/2000
Posts: 12332
Loc: Sterling, VA
|
2. WD 1TB, 2TB and 3TB Green drives. Various subversions of them Hm. I had two WD 2TB drives fail in quick succession in my Synology NAS recently. I wonder if there's something to this. We've had threads on the board here discussing hard drive brand failure experiences. Every time I've had these sorts of talks with other people, the only conclusion I can make is that everyone has horror stories about every brand of hard drive. Personally, I've had two Seagate drives fail on me, so I went all WD and I've never had a problem with any of them. For some people it's the opposite. My Synology has never had anything but WD drives - both green and red - and none of them have failed (yet). I don't know about the Drobo, but a problem with nearly all of the small "home sized" NAS units that I have seen, is they simply don't keep the drives cool enough. Fan or no fan, the airflow design is nearly non-existent.
So I would bet on overheating, leading to media errors (aka. "bad sectors"). What would you consider to be too hot for a NAS drive? According to my dashboard my drives are showing temps of 88, 90, 90 and 99 (not sure why one is so much higher). This is with the fan at an inaudible level (from where I'm sitting), but I could adjust it higher.
_________________________
Matt
|
Top
|
|
|
|
#367173 - 06/07/2016 21:35
Re: Drobo HDDs: high failure rate?
[Re: Dignan]
|
carpal tunnel
Registered: 20/12/1999
Posts: 31591
Loc: Seattle, WA
|
the only conclusion I can make is that everyone has horror stories about every brand of hard drive. Agreed. The only way to know for sure is with large statistical sample sizes, not anecdotal evidence from friends. That's why I love the Backblaze reports so dearly. Though I have to say that when I read the Backblaze reports, I feel vindicated about all the hate I've heaped upon Seagate over the years.
|
Top
|
|
|
|
#367174 - 06/07/2016 22:03
Re: Drobo HDDs: high failure rate?
[Re: tfabris]
|
carpal tunnel
Registered: 08/07/1999
Posts: 5546
Loc: Ajijic, Mexico
|
If it's got a fan on the drives, as long as the fan is working, I would think that would be good. Maybe not... I have three hard drives in my sizeable tower case, plus an SSD tucked away by the CD burner. There is a good sized fan blowing directly on the hard drives. But... the guys who built the computer for me stacked the three hard drives one on top of the other, with no more than one or two millimeters between them, leaving two empty drive bays on top of them. (I told you it was a sizeable case!) This, I think, was less than ideal, and a month ago when I went to vacuum the dust bunnies out of the motherboard, I moved the drives so there is an empty bay between them. I didn't take any before/after temperature measurements, but I'd think they have to be running cooler now. tanstaafl.
_________________________
"There Ain't No Such Thing As A Free Lunch"
|
Top
|
|
|
|
#367175 - 06/07/2016 22:19
Re: Drobo HDDs: high failure rate?
[Re: tanstaafl.]
|
carpal tunnel
Registered: 20/12/1999
Posts: 31591
Loc: Seattle, WA
|
Agreed that they'll all be running cooler now, especially the middle drive. I specifically recall cooking a middle drive in that kind of layout once before.
But remember you're dealing with a case design and bay spacing which was originally meant for IBM PCs with an old, slow, cool-running HDD in one bay and a 5.5" floppy drive in the other bay. Later, that bay system was mostly intended to combine 1-2 HDDs, a CD-ROM, and a floppy. The design of the spacing of the bays isn't really optimal for a giant stack of modern high speed hard disks which put out a lot of heat. What you did, to give them better spacing, was very wise.
Most NAS units I've seen, though, include enough space in their default design to ventilate the drives and get the fan air moving past them. Though I haven't looked at a Drobo close-up: If the Drobo isn't doing that job sufficiently, then I agree that's a problem.
|
Top
|
|
|
|
#367176 - 06/07/2016 23:20
Re: Drobo HDDs: high failure rate?
[Re: tfabris]
|
carpal tunnel
Registered: 18/06/2001
Posts: 2504
Loc: Roma, Italy
|
Did a fan go bad in your Drobo? No... Why?
_________________________
= Taym = MK2a #040103216 * 100Gb *All/Colors* Radio * 3.0a11 * Hijack = taympeg
|
Top
|
|
|
|
#367177 - 07/07/2016 03:23
Re: Drobo HDDs: high failure rate?
[Re: tfabris]
|
carpal tunnel
Registered: 08/03/2000
Posts: 12332
Loc: Sterling, VA
|
That's why I love the Backblaze reports so dearly.
Though I have to say that when I read the Backblaze reports, I feel vindicated about all the hate I've heaped upon Seagate over the years. I haven't seen those reports. I recall that Google released a report a while back, but decided not to name names.
_________________________
Matt
|
Top
|
|
|
|
#367178 - 07/07/2016 03:26
Re: Drobo HDDs: high failure rate?
[Re: Dignan]
|
carpal tunnel
Registered: 20/12/1999
Posts: 31591
Loc: Seattle, WA
|
I haven't seen those reports. I linked the most recent Backblaze report in my first reply in this thread. Here it is again: https://www.backblaze.com/blog/hard-drive-reliability-stats-q1-2016/Interesting things about the reports: Despite Seagate having (until this year) the worst reliability, Backblaze preferred to use them because they were cheap, plentiful, and easy to source. I find this year's report interesting because of the massive improvement Seagate made since 2015. Still, HGST is the king, and has been for quite a while now.
|
Top
|
|
|
|
#367179 - 07/07/2016 10:01
Re: Drobo HDDs: high failure rate?
[Re: Dignan]
|
carpal tunnel
Registered: 18/01/2000
Posts: 5683
Loc: London, UK
|
I bought my Synology two years ago this week, and I've never had a drive fail on me (*knocking on wood*). My Synology has two drives in it. I just checked and they're reporting ~35 °C (95 °F). It's a relatively warm day today (for the UK), and the NAS is in a cupboard. OTOH, one of the disks is reporting a SMART error; it's done that since I first put it in the Synology. Earlier versions of DSM tried spinning all the disks up at once -- these disks (Samsung) wanted more power than it could provide -- disk now registers a permanent spin-up failure. I'm just keeping my fingers crossed that there are no other SMART errors being hidden by this one.
_________________________
-- roger
|
Top
|
|
|
|
#367180 - 07/07/2016 11:44
Re: Drobo HDDs: high failure rate?
[Re: tfabris]
|
pooh-bah
Registered: 12/01/2002
Posts: 2009
Loc: Brisbane, Australia
|
Interesting things about the reports: Despite Seagate having (until this year) the worst reliability, Backblaze preferred to use them because they were cheap, plentiful, and easy to source.
And that's something you need to keep in mind when using their data. They can handle multiple drive failures without issue whereas the average home user can usually tolerate one failure at a time at most. So they value price and availability over reliability as that suits their installation. A home or even small business user would likely not.
_________________________
Christian #40104192 120Gb (no longer in my E36 M3, won't fit the E46 M3)
|
Top
|
|
|
|
#367181 - 07/07/2016 11:49
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 29/08/2000
Posts: 14489
Loc: Canada
|
There's a lot of speculation here, but no need for much of it: the drive firmware records exactly what the issues are, and reports it via the S.M.A.R.T. interface.
Under Linux, one can read all of the data, along with the firmware's specified pain/temperature thresholds, with this command:
smartctl -a /dev/sd?
If a drive got too hot, it will show it. If a drive developed media errors, it will (normally) show it. Except WD drives are known to cheat here.
If the failures are due to bad cabling or loose SATA connectors, it will show it. That happened to our server here just last week: smart data indicated UDMA_CRC errors, so I opened up the box, and could see exposed gold fingers on the connector. I shoved the connector back into place and made a mental note to get out the hot-melt glue gun later on (perhaps today!).
Cheers
Edited by mlord (07/07/2016 11:52) Edit Reason: Added note about cable issues
|
Top
|
|
|
|
#367182 - 07/07/2016 12:01
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 29/08/2000
Posts: 14489
Loc: Canada
|
For those who may never have seen all of this data before, here is the output from the "failed" drive in my server. Nothing wrong with it other than a loose connector at the time. But the drive dutifully logs and remembers such issues. Here is a short excerpt from the data, with only the parts that are most important (to this specific drive) to keep an eye on: Device Model: Crucial_CT960M500SSD1
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE
..
5 Reallocate_NAND_Blk_Cnt 0x0033 100 100 000 Pre-fail Always - 0
171 Program_Fail_Count 0x0032 100 100 000 Old_age Always - 0
172 Erase_Fail_Count 0x0032 100 100 000 Old_age Always - 0
173 Ave_Block-Erase_Count 0x0032 100 100 000 Old_age Always - 8
187 Reported_Uncorrect 0x0032 100 100 000 Old_age Always - 0
194 Temperature_Celsius 0x0022 059 049 000 Old_age Always - 41 (0 51 255 150 0)
196 Reallocated_Event_Count 0x0032 100 100 000 Old_age Always - 0
197 Current_Pending_Sector 0x0032 100 100 000 Old_age Always - 0
198 Offline_Uncorrectable 0x0030 100 100 000 Old_age Offline - 0
199 UDMA_CRC_Error_Count 0x0032 100 100 000 Old_age Always - 314
202 Percent_Lifetime_Used 0x0031 100 100 000 Pre-fail Offline - 0
210 Success_RAIN_Recov_Cnt 0x0032 100 100 000 Old_age Always - 4
The important number here is usually the "RAW_VALUE" field. You can see that the cabling has suffered 314 data transfer errors, which is how I knew to look for a loose connector in the system. Temperatures have ranged as high as 51C for this drive, currently at 41C. The full report now follows: smartctl 6.6 2016-05-31 r4324 [x86_64-linux-4.6.3] (local build)
Copyright (C) 2002-16, Bruce Allen, Christian Franke, www.smartmontools.org
=== START OF INFORMATION SECTION ===
Model Family: Crucial/Micron MX100/MX200/M5x0/M600 Client SSDs
Device Model: Crucial_CT960M500SSD1
Serial Number: xxxxxxxxxxxx
LU WWN Device Id: 5 00a075 xxxxxxxxxx
Firmware Version: MU05
User Capacity: 960,197,124,096 bytes [960 GB]
Sector Sizes: 512 bytes logical, 4096 bytes physical
Rotation Rate: Solid State Device
Form Factor: 2.5 inches
Device is: In smartctl database [for details use: -P show]
ATA Version is: ACS-2, ATA8-ACS T13/1699-D revision 6
SATA Version is: SATA 3.1, 6.0 Gb/s (current: 6.0 Gb/s)
Local Time is: Thu Jul 7 08:16:33 2016 EDT
SMART support is: Available - device has SMART capability.
SMART support is: Enabled
=== START OF READ SMART DATA SECTION ===
SMART overall-health self-assessment test result: PASSED
General SMART Values:
Offline data collection status: (0x82) Offline data collection activity
was completed without error.
Auto Offline Data Collection: Enabled.
Self-test execution status: ( 0) The previous self-test routine completed
without error or no self-test has ever
been run.
Total time to complete Offline
data collection: (10235) seconds.
Offline data collection
capabilities: (0x7b) SMART execute Offline immediate.
Auto Offline data collection on/off support.
Suspend Offline collection upon new
command.
Offline surface scan supported.
Self-test supported.
Conveyance Self-test supported.
Selective Self-test supported.
SMART capabilities: (0x0003) Saves SMART data before entering
power-saving mode.
Supports SMART auto save timer.
Error logging capability: (0x01) Error logging supported.
General Purpose Logging supported.
Short self-test routine
recommended polling time: ( 2) minutes.
Extended self-test routine
recommended polling time: ( 74) minutes.
Conveyance self-test routine
recommended polling time: ( 3) minutes.
SCT capabilities: (0x0035) SCT Status supported.
SCT Feature Control supported.
SCT Data Table supported.
SMART Attributes Data Structure revision number: 16
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE
1 Raw_Read_Error_Rate 0x002f 100 100 000 Pre-fail Always - 9
5 Reallocate_NAND_Blk_Cnt 0x0033 100 100 000 Pre-fail Always - 0
9 Power_On_Hours 0x0032 100 100 000 Old_age Always - 22705
12 Power_Cycle_Count 0x0032 100 100 000 Old_age Always - 33
171 Program_Fail_Count 0x0032 100 100 000 Old_age Always - 0
172 Erase_Fail_Count 0x0032 100 100 000 Old_age Always - 0
173 Ave_Block-Erase_Count 0x0032 100 100 000 Old_age Always - 8
174 Unexpect_Power_Loss_Ct 0x0032 100 100 000 Old_age Always - 20
180 Unused_Reserve_NAND_Blk 0x0033 000 000 000 Pre-fail Always - 16523
183 SATA_Interfac_Downshift 0x0032 100 100 000 Old_age Always - 0
184 Error_Correction_Count 0x0032 100 100 000 Old_age Always - 0
187 Reported_Uncorrect 0x0032 100 100 000 Old_age Always - 0
194 Temperature_Celsius 0x0022 059 049 000 Old_age Always - 41 (0 51 255 150 0)
196 Reallocated_Event_Count 0x0032 100 100 000 Old_age Always - 0
197 Current_Pending_Sector 0x0032 100 100 000 Old_age Always - 0
198 Offline_Uncorrectable 0x0030 100 100 000 Old_age Offline - 0
199 UDMA_CRC_Error_Count 0x0032 100 100 000 Old_age Always - 314
202 Percent_Lifetime_Used 0x0031 100 100 000 Pre-fail Offline - 0
206 Write_Error_Rate 0x000e 100 100 000 Old_age Always - 0
210 Success_RAIN_Recov_Cnt 0x0032 100 100 000 Old_age Always - 4
246 Total_Host_Sector_Write 0x0032 100 100 --- Old_age Always - 7996121592
247 Host_Program_Page_Count 0x0032 100 100 --- Old_age Always - 250001071
248 Bckgnd_Program_Page_Cnt 0x0032 100 100 --- Old_age Always - 363762998
SMART Error Log Version: 1
No Errors Logged
SMART Self-test log structure revision number 1
Num Test_Description Status Remaining LifeTime(hours) LBA_of_first_error
# 1 Vendor (0xff) Completed without error 00% 22696 -
# 2 Vendor (0xff) Completed without error 00% 22682 -
# 3 Vendor (0xff) Completed without error 00% 22668 -
# 4 Vendor (0xff) Completed without error 00% 22654 -
# 5 Vendor (0xff) Completed without error 00% 22640 -
# 6 Vendor (0xff) Completed without error 00% 22626 -
# 7 Vendor (0xff) Completed without error 00% 22613 -
# 8 Vendor (0xff) Completed without error 00% 22599 -
# 9 Vendor (0xff) Completed without error 00% 22585 -
#10 Vendor (0xff) Completed without error 00% 22571 -
#11 Vendor (0xff) Completed without error 00% 22557 -
#12 Vendor (0xff) Completed without error 00% 22543 -
#13 Vendor (0xff) Completed without error 00% 22530 -
#14 Vendor (0xff) Completed without error 00% 22516 -
#15 Vendor (0xff) Completed without error 00% 22502 -
#16 Vendor (0xff) Completed without error 00% 22488 -
#17 Vendor (0xff) Completed without error 00% 22474 -
#18 Vendor (0xff) Completed without error 00% 22460 -
#19 Vendor (0xff) Completed without error 00% 22446 -
#20 Vendor (0xff) Completed without error 00% 22434 -
#21 Vendor (0xff) Completed without error 00% 22420 -
SMART Selective self-test log data structure revision number 1
SPAN MIN_LBA MAX_LBA CURRENT_TEST_STATUS
1 0 0 Not_testing
2 0 0 Not_testing
3 0 0 Not_testing
4 0 0 Not_testing
5 0 0 Not_testing
Selective self-test flags (0x0):
After scanning selected spans, do NOT read-scan remainder of disk.
If Selective self-test is pending on power-up, resume after 0 minute delay.
|
Top
|
|
|
|
#367183 - 07/07/2016 12:27
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 29/08/2000
Posts: 14489
Loc: Canada
|
Oddly enough, I have two of these supposedly identical drives in that system, yet smartctl reports temperatures differently for the second one, with somewhat more readable output:
194 Temperature_Celsius ... 40 (Min/Max 0/46)
|
Top
|
|
|
|
#367184 - 07/07/2016 16:39
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 08/03/2000
Posts: 12332
Loc: Sterling, VA
|
I challenge your assertion that all home NAS devices don't offer proper airflow to keep the drives cool. My drives are hovering around an average of 93 degrees, which seems to be pretty good from what I've been able to read up on (you didn't answer my question about what you thought was an acceptable range).
And those are temperatures in my rather warm office. Where my NAS sits it's regularly 78 degrees (F). When I open up my Synology's front cover, I can see a fairly generous amount of space around the drives, particularly at the ends. The cover its self is designed to let air in through every edge, with the fan being directly on the other side of the unit. It seems that the hottest drive is the one that's farthest-right and furthest from the direct airflow, so that makes sense and is unfortunate, but it's still well within range, IMO.
*edit*
I just tried upping the fan speed in the Synology dashboard, and now all drives are 29-33C. That's with the fan at "full-speed mode." Previously I had it on "quiet mode," and there's also a "cool mode." Full-speed is a tiny bit too loud for my taste (but really not too bad - I could live with it), so I'll probably back it down to cool mode.
Edited by Dignan (07/07/2016 17:32)
_________________________
Matt
|
Top
|
|
|
|
#367185 - 07/07/2016 16:41
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 08/03/2000
Posts: 12332
Loc: Sterling, VA
|
BTW, if anyone is looking for a simple SMART data reader, Speccy is a nice application. It gives you a good amount of information on your computer.
_________________________
Matt
|
Top
|
|
|
|
#367186 - 07/07/2016 18:20
Re: Drobo HDDs: high failure rate?
[Re: Dignan]
|
carpal tunnel
Registered: 08/07/1999
Posts: 5546
Loc: Ajijic, Mexico
|
BTW, if anyone is looking for a simple SMART data reader, Speccy is a nice application. It gives you a good amount of information on your computer. I found it, downloaded it, installed it, and I really like the simple interface. I'm not S.M.A.R.T. enough to make much sense of what Mark posted, but this output (see attached screenshot) even I can understand. A little bit, anyway. Very nice, Thank you. tanstaafl
Attachments
speccy 01.png (199 downloads)
_________________________
"There Ain't No Such Thing As A Free Lunch"
|
Top
|
|
|
|
#367187 - 07/07/2016 21:21
Re: Drobo HDDs: high failure rate?
[Re: Dignan]
|
carpal tunnel
Registered: 29/08/2000
Posts: 14489
Loc: Canada
|
I challenge your assertion that all home NAS devices.. No such assertion to challenge.
|
Top
|
|
|
|
#367188 - 07/07/2016 21:25
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 08/03/2000
Posts: 12332
Loc: Sterling, VA
|
No problem! It's a helpful program in my work, when I don't always know what's inside the computer I'm working on, and the Windows properties screen has such useless information. CCleaner, also by that developer, is a nice utility for cleaning off junk files, and is great for cleaning up unwanted startup entries. I would leave it on my clients' computers for them to use but I'm worried they'll use the built in registry cleaner.
_________________________
Matt
|
Top
|
|
|
|
#367189 - 07/07/2016 21:28
Re: Drobo HDDs: high failure rate?
[Re: mlord]
|
carpal tunnel
Registered: 08/03/2000
Posts: 12332
Loc: Sterling, VA
|
I challenge your assertion that all home NAS devices.. No such assertion to challenge. ...a problem with nearly all of the small "home sized" NAS units that I have seen, is they simply don't keep the drives cool enough. Fan or no fan, the airflow design is nearly non-existent.
_________________________
Matt
|
Top
|
|
|
|
#367190 - 07/07/2016 21:44
Re: Drobo HDDs: high failure rate?
[Re: Dignan]
|
old hand
Registered: 29/05/2002
Posts: 798
Loc: near Toronto, Ontario, Canada
|
... My drives are hovering around an average of 93 degrees ...
And those are temperatures in my rather warm office. Where my NAS sits it's regularly 78 degrees (F)....
I just tried upping the fan speed in the Synology dashboard, and now all drives are 29-33C... Is that 93 degrees C or F ? How are you measuring the Fahrenheit temperatures?
|
Top
|
|
|
|
#367192 - 08/07/2016 13:34
Re: Drobo HDDs: high failure rate?
[Re: Dignan]
|
carpal tunnel
Registered: 29/08/2000
Posts: 14489
Loc: Canada
|
I challenge your assertion that all home NAS devices.. No such assertion to challenge. ...a problem with nearly all of the small "home sized" NAS units that I have seen, is they simply don't keep the drives cool enough. Fan or no fan, the airflow design is nearly non-existent. Exactly. But on a somewhat related note: what does the drive firmware (S.M.A.R.T.) show as the maximum (ever) recorded temperature for those drives? That's the number you should be most interested in. Drives sitting idle don't get nearly as hot as those under heavy load stress. If that's also still good and low for your drives, then no big worries there. Cheers
Edited by mlord (08/07/2016 13:39)
|
Top
|
|
|
|
#367196 - 08/07/2016 17:15
Re: Drobo HDDs: high failure rate?
[Re: mlord]
|
carpal tunnel
Registered: 08/03/2000
Posts: 12332
Loc: Sterling, VA
|
Is that 93 degrees C or F ?
How are you measuring the Fahrenheit temperatures? Lol, that would be way too hot. No, it's F. Apologies for jumping back and forth between the two. The dashboard reports both, and as someone who doesn't usually use Celsius, it's hard for me to stick to those units. Exactly? What does that mean? You say they don't keep the drives cool enough. I say my drives seem pretty cool, but your reaction is...exactly? Huh? what does the drive firmware (S.M.A.R.T.) show as the maximum (ever) recorded temperature for those drives? That's the number you should be most interested in. Drives sitting idle don't get nearly as hot as those under heavy load stress. That's definitely fair. Why didn't you just say that? I've found where Synology reports some SMART info, but it seems incomplete, I don't see the value for historical max temp, and I'm not sure how to read it if I could. Also, I'd argue a little about taking max temp as the most important number here. That tells me that one time the temperature reached a certain level, but that's an incomplete story. I certainly want to avoid whatever that max temp is, but what if it only ever got that hot once and never approached it again? What if the AC went out in my house during the summer and I forgot to turn off my NAS? *edit* Oh my god, I've only spent 5 minutes reading up on how to interpret the SMART temperature values, and my head is going to explode. I would have thought that this would be a pretty simple thing, but I'm seeing about four different formulas to convert the "worst" value to an actual temperature, and none of them make sense...
Edited by Dignan (08/07/2016 17:25)
_________________________
Matt
|
Top
|
|
|
|
#367199 - 08/07/2016 19:51
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 29/08/2000
Posts: 14489
Loc: Canada
|
Yeah, every drive vendor seems to have variations on "the standard" for S.M.A.R.T. data. That's why the tools (eg. smartctl on Linux) have databases of known drives to help them interpret the values for us, so we don't have to.
I agree, a single "worst" value isn't as informative as we'd like, but there's no standard for providing more than just that one number from the drive. A nice historical graph would be much better, but no such thing unless the box's software monitors and logs the temperatures over time. Many do, though.
Cheers
|
Top
|
|
|
|
#367202 - 08/07/2016 21:05
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 18/06/2001
Posts: 2504
Loc: Roma, Italy
|
As to the Drobo, I am not sure you can read SMART data of a USB attached Drobo from your PC. So, unless Drobo firmware reads and exposes SMART data it its gui - it doesn't -, there's no way to know.
I think.
Back to my failed 3TB drive: I could read it from my PC, but I only see 750GB, and it hangs repeatedly. I'd say this is indeed fa failed drive.
Edited by Taym (08/07/2016 21:06)
_________________________
= Taym = MK2a #040103216 * 100Gb *All/Colors* Radio * 3.0a11 * Hijack = taympeg
|
Top
|
|
|
|
#367205 - 09/07/2016 16:17
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 08/03/2000
Posts: 12332
Loc: Sterling, VA
|
Dang, sorry to hear that. Wouldn't the SMART data still be readable even if that drive is failing? Have you checked it out to see what the max temp was?
_________________________
Matt
|
Top
|
|
|
|
#367207 - 09/07/2016 20:14
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 18/06/2001
Posts: 2504
Loc: Roma, Italy
|
Matt, actually I haven't. I'll do that, out of curiosity.
_________________________
= Taym = MK2a #040103216 * 100Gb *All/Colors* Radio * 3.0a11 * Hijack = taympeg
|
Top
|
|
|
|
#367208 - 11/07/2016 00:13
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 17/12/2000
Posts: 2665
Loc: Manteca, California
|
A question about drives that fail from prolonged heat...
Is it primarily the electronics that fail?
_________________________
Glenn
|
Top
|
|
|
|
#367209 - 12/07/2016 12:53
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 29/08/2000
Posts: 14489
Loc: Canada
|
In my experience, it is the media that fails. Dunno if the coatings just become more fragile when HOT, or the platters expand beyond tolerance, or the lubricant escapes, or what.
The electronics on the whole still seem to function after cooling down on such drives, but with an accumulation of "media errors" (aka. "bad sectors").
|
Top
|
|
|
|
#367241 - 25/07/2016 13:29
Re: Drobo HDDs: high failure rate?
[Re: tfabris]
|
carpal tunnel
Registered: 08/07/1999
Posts: 5546
Loc: Ajijic, Mexico
|
The design of the spacing of the bays isn't really optimal for a giant stack of modern high speed hard disks which put out a lot of heat. Yeah... but there are workarounds for that. tanstaafl.
Attachments
_________________________
"There Ain't No Such Thing As A Free Lunch"
|
Top
|
|
|
|
#367247 - 26/07/2016 22:50
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 18/06/2001
Posts: 2504
Loc: Roma, Italy
|
_________________________
= Taym = MK2a #040103216 * 100Gb *All/Colors* Radio * 3.0a11 * Hijack = taympeg
|
Top
|
|
|
|
#367781 - 25/10/2016 23:38
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 18/06/2001
Posts: 2504
Loc: Roma, Italy
|
... and one more failed drive in my Drobo. A WD Green, 4TB. This is starting to get a bit unreal. I am checking the failed drives SMART status tomorrow in the office.
I am now thinking the Drobo is killing them (I would assume poor coolingt). So far the ones I checked were in face dead, so it seems Drobo is correct in that.
I am now keeping my Drobo without the front cover, in the hope cooling improves, IF that is the issue.
Edited by Taym (25/10/2016 23:45)
_________________________
= Taym = MK2a #040103216 * 100Gb *All/Colors* Radio * 3.0a11 * Hijack = taympeg
|
Top
|
|
|
|
#367782 - 26/10/2016 01:05
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 08/03/2000
Posts: 12332
Loc: Sterling, VA
|
That is crazy. I have to think that there's something unusual about your Drobo. I can't imagine that failure rate wouldn't go unnoticed by the company and other Drobo users.
_________________________
Matt
|
Top
|
|
|
|
#367783 - 26/10/2016 09:15
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 18/06/2001
Posts: 2504
Loc: Roma, Italy
|
Matt, I agree.
However, apparently the "broken" HDD is in fact working. I did not manage to get the SMART data from it, as I've only had the time to test if from a USB dock that does not seem to expose those (tried with various HDD test tools), but I could see the drive, make it GPT, create volume, format it, save an read files from it.
Please, notice this is the first time this happened. Of the previous failed drives, two I did check and they had actually failed, others I did not check and simply relied on Drobo telling me they were bad, and then this one.
No obvious pattern to me...
_________________________
= Taym = MK2a #040103216 * 100Gb *All/Colors* Radio * 3.0a11 * Hijack = taympeg
|
Top
|
|
|
|
#367784 - 26/10/2016 12:28
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 29/08/2000
Posts: 14489
Loc: Canada
|
Need to look at the SMART data to see what the drive thinks about it all.
|
Top
|
|
|
|
#367785 - 26/10/2016 22:38
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
old hand
Registered: 29/05/2002
Posts: 798
Loc: near Toronto, Ontario, Canada
|
What about hardware problems in the Drobo electronics? Flakey power supply, crazy controller, intermittent connectors/cables.
Vibration from the supporting shelf or a nearby cooling fan.
Heat is not the only way to hurt a hard drive.
|
Top
|
|
|
|
#367789 - 30/10/2016 12:32
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 18/06/2001
Posts: 2504
Loc: Roma, Italy
|
I tried to get the SMART data, but could not via 3 different USB-SATA adapters, and one eSATA. As soon as I have a minute I am opening one of my PCs case and get it.
K447, thanks for hinting at all other possibile causes. Let's see - Drobo electronics: no way for me to easily find out. I've been planning to replace the fan with some quieter one for a while, so I will eventually have to open the Drobo. I'll inspect it for anything suspicious.
- Drobo does vibrate itself at times. Apparently, many do, and by pushing the HDD firmly in (as recommended by Drobo themselves) it does help: vibration noise drops. Drobo is sitting on a Desk with my home PC, which is mostly off. it is connected to my home server, on a desk nearby, which is on 24H, but vibration from it do not reach where the Drobo is.
- a 650 VA APC UPS with a battery still in good shape is protecting both the drobo and the server.
Edited by Taym (30/10/2016 12:35)
_________________________
= Taym = MK2a #040103216 * 100Gb *All/Colors* Radio * 3.0a11 * Hijack = taympeg
|
Top
|
|
|
|
#367790 - 30/10/2016 13:23
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
old hand
Registered: 29/05/2002
Posts: 798
Loc: near Toronto, Ontario, Canada
|
Intermittent and long cycle time failures can be difficult to diagnose, often impossible to be certain of the cause(s).
You may want to consider replacing this Drobo unit with a different unit or a different model. Retire or sell this one, as you deem appropriate.
The ongoing downside risks with the current Drobo in terms of your time and potential data loss probably outweighs the money value of the Drobo unit, not to mention the cost of replacing hard drives multiple times.
Sometimes it is best to just move on, leave the troublesome unit in the past.
Edited by K447 (30/10/2016 15:11)
|
Top
|
|
|
|
#367791 - 30/10/2016 13:32
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 18/06/2001
Posts: 2504
Loc: Roma, Italy
|
This unit is devoted entirely to local backup sessions, so risk of data loss *EVEN* in worst case scenario nearly none. Otherwise, I would have moved to a Synology already, because I essentially agree with u AND Synology units are so much nicer in terms of build quality and features...
_________________________
= Taym = MK2a #040103216 * 100Gb *All/Colors* Radio * 3.0a11 * Hijack = taympeg
|
Top
|
|
|
|
#367792 - 30/10/2016 23:16
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 08/03/2000
Posts: 12332
Loc: Sterling, VA
|
What were you using to get the SMART data? Have you tried Speccy?
Edited by Dignan (30/10/2016 23:17)
_________________________
Matt
|
Top
|
|
|
|
#367793 - 01/11/2016 13:12
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 18/06/2001
Posts: 2504
Loc: Roma, Italy
|
Yes, i tried Speccy too. Apparently, SMART data is not going through the SATA->USB adapters I am using (all built into three different HDD Docks I normally use at home and in the office). I realize there are some USB adapters+drivers that can pull SMART data from HDDs (or, so I read); apparently not mine.
Not a big problem, I just need to find few minutes to open up a PC case.
I also used: Crystal Disk Info and HDD Scan.
Edited by Taym (01/11/2016 13:13)
_________________________
= Taym = MK2a #040103216 * 100Gb *All/Colors* Radio * 3.0a11 * Hijack = taympeg
|
Top
|
|
|
|
#367794 - 01/11/2016 18:05
Re: Drobo HDDs: high failure rate?
[Re: Taym]
|
carpal tunnel
Registered: 08/03/2000
Posts: 12332
Loc: Sterling, VA
|
Ah sure. Those SATA>USB adapters can be extremely annoying. I've had really bad problems with a couple of the USB3 variety and I'm not sure why. I've used a couple that have prevented me from migrating to a SSD using Samsung's software. For some reason, the software claims there's no drive connected at all. Then I'll swap out for a USB2 adapter I have and it all works fine. I recently ordered an Anker USB3 adapter, though, so I'll see if that one works better.
I'm very curious to hear what the SMART results are.
_________________________
Matt
|
Top
|
|
|
|
|
|