Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
Disk SMART attribute numbers?
View unanswered posts
View posts from last 24 hours

Goto page 1, 2  Next  
Reply to topic    Gentoo Forums Forum Index Off the Wall
View previous topic :: View next topic  
Author Message
pjp
Administrator
Administrator


Joined: 16 Apr 2002
Posts: 16708

PostPosted: Thu Sep 07, 2017 4:36 am    Post subject: Disk SMART attribute numbers? Reply with quote

Is it expected that the Raw_Read_Error_Rate fluctuate like this?

All info extracted from output of 'smartctl -a'. After the time, only attribute columns 1, 2 and last are shown:
Code:
10:15:07   1 Raw_Read_Error_Rate     81207875
11:18:01   1 Raw_Read_Error_Rate    152682653
12:18:01   1 Raw_Read_Error_Rate      1347498
13:18:01   1 Raw_Read_Error_Rate     73859715
14:18:01   1 Raw_Read_Error_Rate    133400526
15:18:01   1 Raw_Read_Error_Rate    202545450
16:18:01   1 Raw_Read_Error_Rate     46635766
17:18:01   1 Raw_Read_Error_Rate    136429063
18:18:01   1 Raw_Read_Error_Rate    190538390
19:18:01   1 Raw_Read_Error_Rate     30966021
20:18:01   1 Raw_Read_Error_Rate     98628825
21:18:01   1 Raw_Read_Error_Rate    172106649
22:18:01   1 Raw_Read_Error_Rate    233733264


I also thought it was interesting then that the Seek_Error_Rate only increases:
Code:
10:15:07   7 Seek_Error_Rate        791524967
11:18:01   7 Seek_Error_Rate        791609724
12:18:01   7 Seek_Error_Rate        791703179
13:18:01   7 Seek_Error_Rate        791784513
14:18:01   7 Seek_Error_Rate        791862186
15:18:01   7 Seek_Error_Rate        791942531
16:18:01   7 Seek_Error_Rate        792031235
17:18:01   7 Seek_Error_Rate        792133871
18:18:01   7 Seek_Error_Rate        792208429
19:18:01   7 Seek_Error_Rate        792297211
20:18:01   7 Seek_Error_Rate        792365714
21:18:01   7 Seek_Error_Rate        792447729
22:18:01   7 Seek_Error_Rate        792529171


The other attributes are 0 or don't change (except for the obvious, like power on count, which is ~10.7 years):
Code:
smartctl 6.4 2015-06-04 r4109 [x86_64-linux-4.12.5-gentoo] (local build)
Copyright (C) 2002-15, Bruce Allen, Christian Franke, www.smartmontools.org

=== START OF READ SMART DATA SECTION ===
SMART Attributes Data Structure revision number: 10
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE      UPDATED  WHEN_FAILED RAW_VALUE
  1 Raw_Read_Error_Rate     0x000f   058   049   006    Pre-fail  Always       -       3638621
  3 Spin_Up_Time            0x0003   098   096   000    Pre-fail  Always       -       0
  4 Start_Stop_Count        0x0032   100   100   020    Old_age   Always       -       91
  5 Reallocated_Sector_Ct   0x0033   100   100   036    Pre-fail  Always       -       23
  7 Seek_Error_Rate         0x000f   088   060   030    Pre-fail  Always       -       792550975
  9 Power_On_Hours          0x0032   001   001   000    Old_age   Always       -       93713
 10 Spin_Retry_Count        0x0013   100   100   097    Pre-fail  Always       -       0
 12 Power_Cycle_Count       0x0032   100   100   020    Old_age   Always       -       100
194 Temperature_Celsius     0x0022   028   047   000    Old_age   Always       -       28 (0 15 0 0 0)
195 Hardware_ECC_Recovered  0x001a   058   048   000    Old_age   Always       -       3638621
197 Current_Pending_Sector  0x0012   100   100   000    Old_age   Always       -       0
198 Offline_Uncorrectable   0x0010   100   100   000    Old_age   Offline      -       0
199 UDMA_CRC_Error_Count    0x003e   200   200   000    Old_age   Always       -       0
200 Multi_Zone_Error_Rate   0x0000   100   253   000    Old_age   Offline      -       0
202 Data_Address_Mark_Errs  0x0032   100   253   000    Old_age   Always       -       0


Hardware_ECC_Recovered fluctuates as well:
Code:
10:15:07 195 Hardware_ECC_Recovered  81207875
11:18:01 195 Hardware_ECC_Recovered 152682653
12:18:01 195 Hardware_ECC_Recovered   1347498
13:18:01 195 Hardware_ECC_Recovered  73859715
14:18:01 195 Hardware_ECC_Recovered 133400526
15:18:01 195 Hardware_ECC_Recovered 202545450
16:18:01 195 Hardware_ECC_Recovered  46635766
17:18:01 195 Hardware_ECC_Recovered 136429063
18:18:01 195 Hardware_ECC_Recovered 190538390
19:18:01 195 Hardware_ECC_Recovered  30966021
20:18:01 195 Hardware_ECC_Recovered  98628825
21:18:01 195 Hardware_ECC_Recovered 172106649
22:18:01 195 Hardware_ECC_Recovered 233733264

_________________
Ladies and gentlemen, allow me to present, Man Not Caring. -- Perry Cox
Back to top
View user's profile Send private message
R0b0t1
Apprentice
Apprentice


Joined: 05 Jun 2008
Posts: 214

PostPosted: Thu Sep 07, 2017 6:23 am    Post subject: Reply with quote

The only obvious thing I can suggest is plotting the failure rates against drive temperature. It might also be due to the control logic used in the disk. If they are using a PID controller then oscillations might be due to bad control constants.

It could also just be some artifact of the drive controller's implementation which we will never know about.
Back to top
View user's profile Send private message
erm67
Apprentice
Apprentice


Joined: 01 Nov 2005
Posts: 184
Location: somewhere in Renziland.

PostPosted: Thu Sep 07, 2017 12:14 pm    Post subject: Reply with quote

I bet it is a disk >2Tb :-)

Anyway 792529171 and 233733264 are RAW values, if smartmontools cannot interpret the raw value correctly (often) it show strange numbers apparently meaningless, you should check with th HD maker to interpret correctly the values.


There are 23 Reallocated_Sector_Ct the disk is definitely damaged but it is using it's spare sectors to reallocate the damaged ones. It might be that every time the OS attempts to acces a bad sector it generates a read/seek error but the error is recovered using a spare sector, however the read/seek error count increases.
in short if the reallocated sector count doesn't change don't worry
if it increases plan for a new disk
if some Uncorrectable appears buy a new disk.

Once upon a time my laptop fell and some errors like that appeared on the disk, some uncorrectable too, I was able to find out where the damaged sectors where ( between the first 16Gb and 100Gb) and I was able to create a partition that contained the damaged sector to isolate them. I used the rest of the disk without problems for several years.
_________________
True ignorance is not the absence of knowledge, but the refusal to acquire it.
A posse ad esse non valet consequentia
Πάντα ῥεῖ
Back to top
View user's profile Send private message
Ant P.
Advocate
Advocate


Joined: 18 Apr 2009
Posts: 4439

PostPosted: Thu Sep 07, 2017 7:13 pm    Post subject: Reply with quote

Those aren't fluctuations, they're wraparounds. Something is going terribly wrong with that drive.
_________________
*.ebuild // /etc/service/*
Back to top
View user's profile Send private message
patrix_neo
Guru
Guru


Joined: 08 Jan 2004
Posts: 428
Location: The Maldives

PostPosted: Thu Sep 07, 2017 7:45 pm    Post subject: Reply with quote

If I would bet on anything, pjp is trolling us, giving out his hooker numbers. I'm not sure though.

@pjp:
It better not be your new 'top 'puter...I hope not.
What do you have, if so, a 2 week guarantee? With that said, I am on Antsy's side on this. Heat or solar CME
Back to top
View user's profile Send private message
pjp
Administrator
Administrator


Joined: 16 Apr 2002
Posts: 16708

PostPosted: Thu Sep 07, 2017 8:21 pm    Post subject: Reply with quote

R0b0t1 wrote:
The only obvious thing I can suggest is plotting the failure rates against drive temperature. It might also be due to the control logic used in the disk. If they are using a PID controller then oscillations might be due to bad control constants.

It could also just be some artifact of the drive controller's implementation which we will never know about.
I think I vaguely recall seeing some failures a long time ago. At the time, I wasn't too concerned and didn't want to deal with it, so I just ignored it. I started paying attention more recently just because I saw some messages I didn't recognize. That's when I started logging the output.

Thanks.


erm67 wrote:
I bet it is a disk >2Tb :-)
Actually, only 1TB. Are problems common on >2TB drives beyond the liklihood that with more sectors, the chance increases to have a bad one? (EDIT: LOL it isn't even that... I think this is an ancient 120GB drive... from fdisk "Disk /dev/sda: 111.8 GiB" --- see my comment in the next post about having forgot what hardware was in this thing).

erm67 wrote:
Anyway 792529171 and 233733264 are RAW values, if smartmontools cannot interpret the raw value correctly (often) it show strange numbers apparently meaningless, you should check with th HD maker to interpret correctly the values.
Wikipedia indicates that as a vendor specific metric, it isn't particularly useful by itself. I'm not sure its worth digging through the mfg docs just yet. I was mainly curious if there were any generalized impressions. Mainly because I expected an increase for the number, not up and down.

erm67 wrote:
There are 23 Reallocated_Sector_Ct the disk is definitely damaged but it is using it's spare sectors to reallocate the damaged ones. It might be that every time the OS attempts to acces a bad sector it generates a read/seek error but the error is recovered using a spare sector, however the read/seek error count increases.
in short if the reallocated sector count doesn't change don't worry
if it increases plan for a new disk
if some Uncorrectable appears buy a new disk.

Once upon a time my laptop fell and some errors like that appeared on the disk, some uncorrectable too, I was able to find out where the damaged sectors where ( between the first 16Gb and 100Gb) and I was able to create a partition that contained the damaged sector to isolate them. I used the rest of the disk without problems for several years.
Monitoring is the short term goal. I powered on another system to get it ready for taking over. It's also old, but not as old.

Thanks.
_________________
Ladies and gentlemen, allow me to present, Man Not Caring. -- Perry Cox


Last edited by pjp on Thu Sep 07, 2017 8:43 pm; edited 1 time in total
Back to top
View user's profile Send private message
pjp
Administrator
Administrator


Joined: 16 Apr 2002
Posts: 16708

PostPosted: Thu Sep 07, 2017 8:27 pm    Post subject: Reply with quote

Ant P. wrote:
Those aren't fluctuations, they're wraparounds. Something is going terribly wrong with that drive.
I wondered if that was possible. Didn't sound good when I thought of it :). So I decided to make an escape option more of a priority than it had been.


patrix_neo wrote:
If I would bet on anything, pjp is trolling us, giving out his hooker numbers. I'm not sure though.

@pjp:
It better not be your new 'top 'puter...I hope not.
What do you have, if so, a 2 week guarantee? With that said, I am on Antsy's side on this. Heat or solar CME
New? No. It's so old I forgot what I had in it. I thought it was an ASUS mobo with 16GB of RAM. Turns out it is a Gigabyte with only 4GB of RAM. The other compute I have has only 16GB. Could've sworn I had 32GB. I must have decided against that since it wasn't ECC.
_________________
Ladies and gentlemen, allow me to present, Man Not Caring. -- Perry Cox
Back to top
View user's profile Send private message
patrix_neo
Guru
Guru


Joined: 08 Jan 2004
Posts: 428
Location: The Maldives

PostPosted: Thu Sep 07, 2017 9:12 pm    Post subject: Reply with quote

pjp wrote:
Ant P. wrote:
Those aren't fluctuations, they're wraparounds. Something is going terribly wrong with that drive.
I wondered if that was possible. Didn't sound good when I thought of it :). So I decided to make an escape option more of a priority than it had been.


patrix_neo wrote:
{stuff}
New? No. It's so old I forgot what I had in it. I thought it was an ASUS mobo with 16GB of RAM. Turns out it is a Gigabyte with only 4GB of RAM. The other compute I have has only 16GB. Could've sworn I had 32GB. I must have decided against that since it wasn't ECC.


You like to flail around with fancy words, don't you? (MB/GB) Have you heard of the new kids on the block? MiB and GiB?
And, you better check you're not running Windows 98.

I was actually wondering, because of that thread of yours about investing in a newer one.
Back to top
View user's profile Send private message
pjp
Administrator
Administrator


Joined: 16 Apr 2002
Posts: 16708

PostPosted: Thu Sep 07, 2017 11:13 pm    Post subject: Reply with quote

patrix_neo wrote:
You like to flail around with fancy words, don't you? (MB/GB) Have you heard of the new kids on the block? MiB and GiB?
I want a generic version, "gig", "meg". The details aren't important most of the time. The "ibi" stuff is about one of the most idiotic naming conventions I ever encountered. Plenty of stuff just uses "G" or "M".

patrix_neo wrote:
I was actually wondering, because of that thread of yours about investing in a newer one.
Recently? I don't think I've given consideration to an upgrade since the last one. Oh, you probably mean my laptop. For the time being, that's still Windows 10. Which reminds me every time I do updates that I wish *nix was better supported on mobile hardware.
_________________
Ladies and gentlemen, allow me to present, Man Not Caring. -- Perry Cox
Back to top
View user's profile Send private message
Bones McCracker
Veteran
Veteran


Joined: 14 Mar 2006
Posts: 1605
Location: U.S.A.

PostPosted: Fri Sep 08, 2017 12:32 am    Post subject: Reply with quote

The only thing spinning disks still belong in is RAID arrays and logical volumes providing appropriate redundancy. Get a SSD.
Back to top
View user's profile Send private message
pjp
Administrator
Administrator


Joined: 16 Apr 2002
Posts: 16708

PostPosted: Fri Sep 08, 2017 2:39 am    Post subject: Reply with quote

1TB SSD are still too expensive and just not worth it for my needs. Getting something smaller doesn't eliminate the need for spinning platters, so minimal gain.
_________________
Ladies and gentlemen, allow me to present, Man Not Caring. -- Perry Cox
Back to top
View user's profile Send private message
erm67
Apprentice
Apprentice


Joined: 01 Nov 2005
Posts: 184
Location: somewhere in Renziland.

PostPosted: Fri Sep 08, 2017 7:01 am    Post subject: Reply with quote

I am totally satisfied with my home NAS, I used the 2 2.5" 1Tb rotational disks that came with our laptops and bought 2 SSD, It is better to buy a laptop with a rotational and put an SSD in it, it's cheaper and the SSD probably better. To work confortably with the NAS a gbit cable is required but wi-fi is getting faster quickly, at 270Mbit it is enjoyable if I am alone.
People look strangely when I say NAS with 2.5" but actually it is an excellent idea, It is very quiet and requires little power. Using 2.5" disks datacenters can pack lots of HD in a 1U and power consumption is also a concern for them, in fact there are (very expensive) HiEnd 2.5" disks specially designed for NAS and datacenters ...
It is just a matter of physics, the energy required to spin and operate a 3.5" splatter is a lot higher, so it is the stress on the mechanical components, A smaller disk of comparable quality will last longer. There are very little reasons to buy 3.5" disk nowadays ....
_________________
True ignorance is not the absence of knowledge, but the refusal to acquire it.
A posse ad esse non valet consequentia
Πάντα ῥεῖ
Back to top
View user's profile Send private message
pjp
Administrator
Administrator


Joined: 16 Apr 2002
Posts: 16708

PostPosted: Fri Sep 08, 2017 3:01 pm    Post subject: Reply with quote

2.5" definitely, though I don't know how laptop drives compare. I suppose at least as good as other consumer drives, possibly better due to the mobile needs. Only on older hardware have I seen enterprise servers with 3.5" disks.

Are you just using a white box solution for the NAS, or something pre-built?

I have a platter drive in my laptop currently, and plan to try getting Linux and/or BSD working on it. I hadn't considered repurposing it for a NAS. My main complaint about white box solutions is the difficulty in identifying which physical drive is the failure (yes, I've bee spoiled by lights automatically turning on when a failure occurs).
_________________
Ladies and gentlemen, allow me to present, Man Not Caring. -- Perry Cox
Back to top
View user's profile Send private message
erm67
Apprentice
Apprentice


Joined: 01 Nov 2005
Posts: 184
Location: somewhere in Renziland.

PostPosted: Fri Sep 08, 2017 5:51 pm    Post subject: Reply with quote

Well actually when I was looking for a new laptop some years ago, I noticed that it was considerably cheaper buy an otherwise identical laptop with a rotational disk instead of an ssd (or hybrid disk), since I already had 2 nice SSD I had bought to refurbish our old lappys, I bought 2 laptops with a 1Tb rotational disk on board and imediately voided the warranty replacing the disk (actually it can be argued if I did, probably it not). At that point I had two nice 1Tb disks so I got a Zyxel NSA-325 on ebay for a few € (it was already out of production), put archlinux arm and the 2 1Tb platters on it.
It's really great as a NAS .... Sometimes I fancy to upgrade with something better (with dual Gbit interface with link aggregation for example) or with more memory but otherwise I am satisfied.
It only has 512Mb ram so it can't be a server, only a NAS, no gentoo either (too little ram, and not a great idea anyway).
If you have some rotational platter at hand try archlinuxarm.org they support a lot of NAS box that can be bought inexpensively, or if you're rich get a qnap or a synology box :-)
_________________
True ignorance is not the absence of knowledge, but the refusal to acquire it.
A posse ad esse non valet consequentia
Πάντα ῥεῖ
Back to top
View user's profile Send private message
pjp
Administrator
Administrator


Joined: 16 Apr 2002
Posts: 16708

PostPosted: Fri Sep 08, 2017 10:06 pm    Post subject: Reply with quote

I'm surprised enough people buy home quality 2-bay chassis solutions given the prices. You'd think the components were rare or something. There just high enough that it seems worth a custom build.
_________________
Ladies and gentlemen, allow me to present, Man Not Caring. -- Perry Cox
Back to top
View user's profile Send private message
Bones McCracker
Veteran
Veteran


Joined: 14 Mar 2006
Posts: 1605
Location: U.S.A.

PostPosted: Sat Sep 09, 2017 3:58 pm    Post subject: Reply with quote

erm67 wrote:
I am totally satisfied with my home NAS, I used the 2 2.5" 1Tb rotational disks that came with our laptops and bought 2 SSD, It is better to buy a laptop with a rotational and put an SSD in it, it's cheaper and the SSD probably better. To work confortably with the NAS a gbit cable is required but wi-fi is getting faster quickly, at 270Mbit it is enjoyable if I am alone.
People look strangely when I say NAS with 2.5" but actually it is an excellent idea, It is very quiet and requires little power. Using 2.5" disks datacenters can pack lots of HD in a 1U and power consumption is also a concern for them, in fact there are (very expensive) HiEnd 2.5" disks specially designed for NAS and datacenters ...
It is just a matter of physics, the energy required to spin and operate a 3.5" splatter is a lot higher, so it is the stress on the mechanical components, A smaller disk of comparable quality will last longer. There are very little reasons to buy 3.5" disk nowadays ....

This is a good approach, in my opinion. NAS provides a redundant array of spinning disks, for cheap storage. Use laptop for end-user OS, for speed.

At this point in time, I would go with 15K SAS, with 3.5 vs 2.5 size driven by backplanes available for the machine). Such higher-end disk technologies are now quite cost-effective, even for home use, and increased LAN speeds have widened the backbone for network storage without resorting to things like infiniband.

One practical, cost-effective approach is to create a wired 1000-Base-T LAN for use exclusively as a storage backbone, while using 802.11x WiFi for general purpose LAN.
Back to top
View user's profile Send private message
erm67
Apprentice
Apprentice


Joined: 01 Nov 2005
Posts: 184
Location: somewhere in Renziland.

PostPosted: Sat Sep 09, 2017 9:27 pm    Post subject: Reply with quote

pjp wrote:
I'm surprised enough people buy home quality 2-bay chassis solutions given the prices. You'd think the components were rare or something. There just high enough that it seems worth a custom build.

If you are talking about the price of Synology, qnap, thecus etc NAS boxes I tend to agree, but the components are not so easy to find and assemble, and not exactly cheap. Still 400 bucks for a 2 bay chassis is a lot of money.
Intel/AMD stuff sucks a lot more power and most boards requires a fan, ARM boards with multiple sata are still expensive and hard to find and BTW are the same boards mounted in those NAS ...
Like boards based on this SoM:
https://www.marvell.com/embedded-processors/armada-38x/
The smallest has 2 SATA ports and 2 gbit port and only requires 3W
boards with SoM like the smaller armada can cost 100$ and more, plus the chassis and power supply, if you want to get serious and buy a MacchiatoBin it's 400$...... (OK it's A LOT better than a 2bay synology box, but still expensive for home usage)

Actually the HW is not important at this point, the disks are faster than network and most arm board can do gbit & SATA full speed without problems, there are other considerations like aestetic, power consumption and noise. I do regularly backup importand stuff from the NAS to a portable disk, so also reliability is relatively important, if the NAS is down I only need ear plugs because the kids can't watch their movies ...
I used mini-ITX boards before, with 3.5 disk, the setup was more expensive than an average chassis and frankly performed the same, only it was more noisy and sucked a lot more power 40-50W compared to 8-10W of the current setup with an arm board and 2 2.5". It makes a big difference over time. It is also almost noiseless, impossible to hear at 1 mt and looks good. The alternative is get a older box supported by a community distro and only use the hardware, like I did I got mine for 50€ on ebay. Just make sure it is possible and safe install a regular linux distro on it.

Maybe once 2-5-10Gbit will be available at home it will be worth buying better hardware, I wait until MacchiatoBin prices go down considerably.
_________________
True ignorance is not the absence of knowledge, but the refusal to acquire it.
A posse ad esse non valet consequentia
Πάντα ῥεῖ
Back to top
View user's profile Send private message
bunder
Bodhisattva
Bodhisattva


Joined: 10 Apr 2004
Posts: 5288

PostPosted: Sun Sep 10, 2017 12:41 am    Post subject: Reply with quote

Mine tend to fluctuate, but the number is usually pretty low.

https://wiki.lime-technology.com/Understanding_SMART_Reports

Quote:
1 Raw_Read_Error_Rate

This is an indicator of the current rate of errors of the low level physical sector read operations. In normal operation, there are ALWAYS a small number of errors when attempting to read sectors, but as long as the number remains small, there is NO issue with the drive. Error correction information and retry mechanisms are in place to catch and fix these errors. Manufacturers therefore determine an optimal level of errors for each drive model, and set up an appropriate scale for monitoring the current error rate. For example, if 3 errors per 1000 read operations seems near perfect to the manufacturer, then an error rate of 3 per 1000 ops might be set to an attribute VALUE of 100. If the rate increased to 10 per 1000, then the rate might be scaled to 80 (completely under manufacturer control, and NEVER revealed or explained to us!).
They are called Raw Reads to distinguish them from the more common term 'read errors', which represent a much higher level read operation. What we usually refer to as a 'read error' is an error returned by a read process, that has attempted a series of one or more seeks and raw reads, plus optional error corrections and retries. It either returns an indicator of total success plus the sector data (considered to be in perfect shape), or it returns an error code, and no sector data.
PLEASE completely ignore the RAW_VALUE number! Only Seagates report the raw value, which yes, does appear to be the number of raw read errors, but should be ignored, completely. All other drives have raw read errors too, but do not report them, leaving this value as zero only. To repeat, Seagates are not worse than other drives because they appear to have raw read errors, rather they are the only one to report the number. I suspect that others do not report the number to avoid a lot of confusion, and questions for their tech support people. Seagate leaves those of us who provide tech support the job of answering the constant questions about this number. Hopefully now that you understand this, you will never bother a kind IT person with questions about the Raw_Read_Error_Rate RAW_VALUE again?
[incomplete?]
Critical attribute - if its WORST falls below its THRESH, then the drive will be considered FAILED


https://en.wikipedia.org/wiki/S.M.A.R.T.#Known_ATA_S.M.A.R.T._attributes

Quote:
(Vendor specific raw value.) Stores data related to the rate of hardware read errors that occurred when reading data from a disk surface. The raw value has different structure for different vendors and is often not meaningful as a decimal number.


I like that Lime tech page, they have an interesting write up on libata errors as well.
Back to top
View user's profile Send private message
pjp
Administrator
Administrator


Joined: 16 Apr 2002
Posts: 16708

PostPosted: Sun Sep 10, 2017 2:19 am    Post subject: Reply with quote

bunder wrote:
Mine tend to fluctuate, but the number is usually pretty low.

[...]

I like that Lime tech page, they have an interesting write up on libata errors as well.
Interesting, thanks. Ultimately it really does seem like the numbers are meaningless without vendor details.

They mention that column 4 is "one of the most important values in the table." And then "You can generally think of it as representing a scale starting at 100% good, then slowly dropping until failure at some predetermined percentage number, in the THRESHOLD column." But when I look at that column for Raw_Read_Error_Rate, it goes up and down slightly. In the short time I've been watching it, the low has been 52 with a high of 59. The WORST and THRESH values haven't changed.

At this point given how many hours it has been on, I'm assuming it's days (or hours) are numbered.
_________________
Ladies and gentlemen, allow me to present, Man Not Caring. -- Perry Cox
Back to top
View user's profile Send private message
pjp
Administrator
Administrator


Joined: 16 Apr 2002
Posts: 16708

PostPosted: Sun Sep 10, 2017 2:27 am    Post subject: Reply with quote

erm67 wrote:
If you are talking about the price of Synology, qnap, thecus etc NAS boxes I tend to agree, but the components are not so easy to find and assemble, and not exactly cheap. Still 400 bucks for a 2 bay chassis is a lot of money.
Intel/AMD stuff sucks a lot more power and most boards requires a fan, ARM boards with multiple sata are still expensive and hard to find and BTW are the same boards mounted in those NAS ...
Like boards based on this SoM:
https://www.marvell.com/embedded-processors/armada-38x/
The smallest has 2 SATA ports and 2 gbit port and only requires 3W
boards with SoM like the smaller armada can cost 100$ and more, plus the chassis and power supply, if you want to get serious and buy a MacchiatoBin it's 400$...... (OK it's A LOT better than a 2bay synology box, but still expensive for home usage)
Yes, stuff like Synology and Qnap. Even HP and WD or whomever makes them. I agree completely with the Intel/AMD part, power consumption and cost. That's why I've never built anything. I think I've asked here a couple of times for low power options, but there just doesn't seem to be much available.

erm67 wrote:
Actually the HW is not important at this point, the disks are faster than network and most arm board can do gbit & SATA full speed without problems, there are other considerations like aestetic, power consumption and noise. I do regularly backup importand stuff from the NAS to a portable disk, so also reliability is relatively important, if the NAS is down I only need ear plugs because the kids can't watch their movies ...
I used mini-ITX boards before, with 3.5 disk, the setup was more expensive than an average chassis and frankly performed the same, only it was more noisy and sucked a lot more power 40-50W compared to 8-10W of the current setup with an arm board and 2 2.5". It makes a big difference over time. It is also almost noiseless, impossible to hear at 1 mt and looks good. The alternative is get a older box supported by a community distro and only use the hardware, like I did I got mine for 50€ on ebay. Just make sure it is possible and safe install a regular linux distro on it.

Maybe once 2-5-10Gbit will be available at home it will be worth buying better hardware, I wait until MacchiatoBin prices go down considerably.
I wish I had the engineering skills and/or the money, because I think there's enough of a market to offer some consumer level options at consumer level pricing. Something with 2-3 drives in a unobtrusive design would probably become invisible. Hooking up a few of those in P2P mode might work really well.
_________________
Ladies and gentlemen, allow me to present, Man Not Caring. -- Perry Cox
Back to top
View user's profile Send private message
erm67
Apprentice
Apprentice


Joined: 01 Nov 2005
Posts: 184
Location: somewhere in Renziland.

PostPosted: Sun Sep 10, 2017 3:37 pm    Post subject: Reply with quote

pjp wrote:
I wish I had the engineering skills and/or the money, because I think there's enough of a market to offer some consumer level options at consumer level pricing. Something with 2-3 drives in a unobtrusive design would probably become invisible. Hooking up a few of those in P2P mode might work really well.


have a look at helios4 it's really tempting, their kickstarter campaign failed but the project apparently goes on and they claim they will deliver the first batch soon. The price &features are OK, but it doesn't look exactly like something I'd put in my living room. The WAF (Wife Acceptance Factor) is probably very low ...
)

Did I mention that marvell armada SoM are supported by freebsd? If it wasn't that it's probably a 32bit CPU and ZFS really needs 64bit it was interesting for a ZFS NAS.... Marvell announced that the next generation of their SoM will be 64bit.

I came to the conclusion that it is not a good idea to run too much software on a NAS (unless it has lots of CPUs & RAM of course) and transmission in particular. That is why I am using an android TV box as a home server.
transmission on the nas will slow it down considerably, probably too many context switches, will also use a lot of non swappable memory for network buffers, it was too heavy for my small box. also the networking part will suffer because of the small packets sent and received by transmission. iscsi in particular really prefers to have a dedicated network interface.
_________________
True ignorance is not the absence of knowledge, but the refusal to acquire it.
A posse ad esse non valet consequentia
Πάντα ῥεῖ
Back to top
View user's profile Send private message
wildhorse
Tux's lil' helper
Tux's lil' helper


Joined: 16 Mar 2006
Posts: 149
Location: Estados Unidos De América

PostPosted: Sun Sep 10, 2017 9:33 pm    Post subject: Reply with quote

I presume that is an error rate which you would get in Florida right now. Just hold on to your disks.

Seriously, did you mount the disks on some soft rubber? If that is the case, then attach them with screws to a heavy frame and see if it makes a difference. Never ever let hard disks vibrate, unless you use them for something else.
Back to top
View user's profile Send private message
R0b0t1
Apprentice
Apprentice


Joined: 05 Jun 2008
Posts: 214

PostPosted: Sun Sep 10, 2017 11:46 pm    Post subject: Reply with quote

pjp wrote:
erm67 wrote:
I bet it is a disk >2Tb :-)
Actually, only 1TB. Are problems common on >2TB drives beyond the liklihood that with more sectors, the chance increases to have a bad one? (EDIT: LOL it isn't even that... I think this is an ancient 120GB drive... from fdisk "Disk /dev/sda: 111.8 GiB" --- see my comment in the next post about having forgot what hardware was in this thing).
That is part of it, but to get densities >=2TB they had to make some compromises. The bits have a greater chance of randomly flipping and manufacturers were, apparently, not able to adequately compensate this.

erm67 wrote:
Anyway 792529171 and 233733264 are RAW values, if smartmontools cannot interpret the raw value correctly (often) it show strange numbers apparently meaningless, you should check with th HD maker to interpret correctly the values.

erm67 wrote:
There are 23 Reallocated_Sector_Ct the disk is definitely damaged but it is using it's spare sectors to reallocate the damaged ones. It might be that every time the OS attempts to acces a bad sector it generates a read/seek error but the error is recovered using a spare sector, however the read/seek error count increases.
in short if the reallocated sector count doesn't change don't worry
if it increases plan for a new disk
if some Uncorrectable appears buy a new disk.

Once upon a time my laptop fell and some errors like that appeared on the disk, some uncorrectable too, I was able to find out where the damaged sectors where ( between the first 16Gb and 100Gb) and I was able to create a partition that contained the damaged sector to isolate them. I used the rest of the disk without problems for several years.
If it is true that the value is wrapping because every (properly handled) read error increments the count, then what good is that field?
Back to top
View user's profile Send private message
erm67
Apprentice
Apprentice


Joined: 01 Nov 2005
Posts: 184
Location: somewhere in Renziland.

PostPosted: Mon Sep 11, 2017 10:25 am    Post subject: Reply with quote

R0b0t1 wrote:
If it is true that the value is wrapping because every (properly handled) read error increments the count, then what good is that field?


It is probably a bitfield that smartmon tool incorrectly convert to an int, probably the correct value is in some of the bits, like the last 6 bits or so. Seagate uses the last 8 bit for the correct value and other bits for transient failures, but the value reported by pjp appear more random. I tried (briefly) to look at the binary representation of the values to see if a pattern was apparent but failed to find it.
The problem is that the value fluctuates, so it is not a total, and it is way to high to be the number of errors in a short time, after an error the controller enter error recovery mode and tries to handle it this takes time, probably respins the platters, parks the heads to zero them and so onif the error is corrected operations go on, otherwise i, if possible, the sector is relocated or marked as uncorrectable. Since the number of relocated sectors do not increase and there aren't uncorrectable sectors, all those errors were fixed by the firmware somehow. We don't know what that number represents.
_________________
True ignorance is not the absence of knowledge, but the refusal to acquire it.
A posse ad esse non valet consequentia
Πάντα ῥεῖ
Back to top
View user's profile Send private message
pjp
Administrator
Administrator


Joined: 16 Apr 2002
Posts: 16708

PostPosted: Mon Sep 11, 2017 5:35 pm    Post subject: Reply with quote

erm67 wrote:
have a look at helios4 it's really tempting, their kickstarter campaign failed but the project apparently goes on and they claim they will deliver the first batch soon. The price &features are OK, but it doesn't look exactly like something I'd put in my living room. The WAF (Wife Acceptance Factor) is probably very low ...
)

Did I mention that marvell armada SoM are supported by freebsd? If it wasn't that it's probably a 32bit CPU and ZFS really needs 64bit it was interesting for a ZFS NAS.... Marvell announced that the next generation of their SoM will be 64bit.
Yeah, 32-bit is a bit disappointing. But that board is small enough to fit into a better case.

erm67 wrote:
I came to the conclusion that it is not a good idea to run too much software on a NAS (unless it has lots of CPUs & RAM of course) and transmission in particular. That is why I am using an android TV box as a home server.
transmission on the nas will slow it down considerably, probably too many context switches, will also use a lot of non swappable memory for network buffers, it was too heavy for my small box. also the networking part will suffer because of the small packets sent and received by transmission. iscsi in particular really prefers to have a dedicated network interface.
Yeah, intensive workloads often prefer to be left alone. Have you tried iSCSI on a consumer network switch?
_________________
Ladies and gentlemen, allow me to present, Man Not Caring. -- Perry Cox
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Off the Wall All times are GMT
Goto page 1, 2  Next
Page 1 of 2

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum