• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

External enclosure problems and Data corruption

lazimutha

n00b
Joined
Jul 29, 2013
Messages
61
Dear all, looking for some insight regarding my problems and recommendations.

I have a coolermaster X-craft enclosure that I have had for a while and use for all of my backups. I use this with a variety of HDDs and didn't really have a problem with it as far as I knew. However, I noticed that a few files were corrupted about 1-2 years ago. Nothing major; a few photos here and there. Not sure if videos were affected, but then again I didn't go through all of them.

Around 6 months ago, my 1TB WD Caviar black in the enclosure had issues with scan disk. It would pop up intermittently and ask me to scan, and after repair it would seem to be alright. At around the same time I started using Teracopy and began to notice CRC errors for some files. Again, not all, and there was only a few kb difference here and there. Then my 1.5TB WD caviar black started playing up. CRC errors to begin with and now problems with scan disk.

I use my laptop as my main computer (asus ul80jt) and multiple HDDs with the coolermaster enclosure. I originally thought that there were memory issues with my laptop, causing these problems, but multiple passes with memtest 86 makes me think that part is alright. Now I have read several accounts of the actual enclosure being a problem. This whole encounter has raised a few questions I would like to ask to you all as I am a relative noobie with regards to data validation etc.:

1. What sort of workflow do you use when trying to identify which part of the computer chain is causing the copy/data corruption problems? I was basically a bit random and am still not completely sure which is the point of failure.

2. Do you think both HDDs are the problem? I have never had an issue with WD and the thought of both of my caviar blacks failing at a similar time, whilst possible, seems rare.

3. If you feel the enclosure is the problem, do you have any recommendations? I am currently looking at the Silverstone TS07.

4. What procedures do you have in place to prevent data corruption? I am keen to secure my files, although I am also aware that I can only devote limited time/resources/brain power for this.

5. What is the best method of pulling all possible (validated) data off of my hard drive? Or, if I just copied it all, what is the best way to check for corruption (I haven't made copies of my media).

Thanks again!
 
Please post a CrystalDiskInfo screenshot of both drives first. The smart data can help identifying whether the drives have issues. I have to say I hate external drive cases with a passion. They usually use cheap controllers and cheap power supplies. You usually have no indicator of the enclosure's PSU starting to fail or getting unreliable. They also do not provide enough airflow for a drive like the WD Black.

From my experience the component responsible for your problem is most likely either the PSU or the internal circuitry of your enclosure or your harddrive. It is much less likely that your laptop has a problem.

If you need multiple HDDs anyway I would like to recommend a proper NAS with an included power supply.

As means against silent data corruption the first recommendation you usually get here is to setup a fileserver with ZFS as filesystem and use a RAIDZ or RAIDZ2 pool with multiple harddrives. That way you are safe against single or double harddrive failures and corruption of the harddrive contents. This does however not protect against cheap hardware. If you have memory corruption usually no filesystem or storage system can help you. Use ECC memory for protection against that. Make backups of all your data!
 
Last edited:
Thanks for that. I can't do the crystaldiskinfo at the moment as I don't want to use the coolermaster enclosure. Although I am planning to buy the Silverstone one unless someone has a better recommendation.

With regards to NAS, my plan was to eventual get a Synology NAS as I have only heard good things about it. However, I read that ZFS wasn't available on that. Also, again from what I have read, the implementation of ZFS is not for the beginner and I ideally would like to buy an all-in-one solution, hence the Synology. Does the Synology come with ECC ram? I have no idea.
 
Thanks Shetu, but I am looking for a home solution. Was thinking of around 8-10GB of space ideally.
 
Please post a CrystalDiskInfo screenshot of both drives first. The smart data can help identifying whether the drives have issues.

This would be my first step. Determine if the drives are actually bad.
 
Went out to buy the Silverstone enclosure so I can run crystaldisk on my hard drives. However, I spent all day walking around and wasn't able to find a shop that stocked them. Instead, there is a Japanese company called 'Century' and another called 'Stardom' (I think that's what it's called. It's definitely Star-something). Does anyone here any information on them? Reliability?

I was also thinking about purchasing a Synology DS413. Is it possible to just use that as a simple enclosure for my ?faulty HDDs, so that I can transfer the data off them, and then work out whether they are actually faulty. Before I then use the NAS to NAS-stuff. May save me money on buying an additional enclosure. Also, if I am able to use the NAS, and I do some file transfers between HDDs within the NAS, would it be routed from HDD to my laptop to HDD, or just HDD to HDD? I am hoping it's the latter as I only have USB2.0 and it would be really really slow.

Thanks.
 
is this a powered enclosure or a USB powered one? if it is powered by an external source. get another PSU to test with
 
Thanks YeuEmMaiMai. It's a powered enclosure. I'll see if I have another PSU to try. What method would you use to test it? I have a spare 360GB drive I know if good so I can use that.
 
So the PSU connector is not something I have a spare of, so I can't check that. In the meantime, I have tried out the WD lifeguard diagnostics on a 160GB disk I have, which passed the extended test. Have also done a bit of checking on the forums and have downloaded PartedMagic. so I'll probably play with that in a bit.

With my current enclosure, here are the crystalmarkinfo details for the two WD drives that have corruption on them:

WD15CDI_zps49a70ba7.jpg


WD10CDI_zpsf9b5000e.jpg


If I run a variety of diagnostic software on these drives, such as badblocks etc., and it passes, does this mean both the HDD and the enclosure are ok? :confused:
 
Just an update. I bought a 2TB WD my passport drive to back up my data. I ran a 'quick test' and a 'write zeros' test using the lifeguard software, which the drive passed. So I thought I would then use the WD acronis software to make clone the ?corrupt 1.5tb caviar black. Acronis asked me if I wanted to 'initialise' the 2TB HD, which I said yes to. After that, neither my laptop, nor a nearby desktop recognised the drive. I could only find it in disk manager, which said I didn't have a partition. So I created that and tried to format, but it said it was being use, so I agreed to 'force format', which then failed.

After trying this a few times, I decided to use partedmagic to try to sort it out, which seemed to work I think. Anyone have any idea what happened and how to stop this happening again? I must have done something noobish, but I have no idea what.

I am currently cloning the 1.5tb drive with clonezilla and fingers crossed it works. If it does, I plan to verify my cloned data before using badblocks to test the hell out of the 1.5tb drive.

I feel as if the more I learn the less I know. I'm sure this sort of thing only happens to ppl who know nothing about backups/data security etc. Ignorance is bliss!

Cheers.
 
Don't buy those prebuilt external drives.

Buy a decent enclosure (like the Coolermaster) with active cooling and eSATA connections and install whatever drive you choose.

It just sounds to me that the drives you were using were going bad and you ignored them to the point where you are now.

There's no way I would try to repair/diagnose a drive thru USB.

At the very least hook the drive thru your MBs SATA ports on your desktop. Never mind, I forgot you are using a laptop with no eSATA.
 
But Old Hippie, I think the coolermaster enclosure was the problem in the first place. The HDD was only used a backup and after it started going weird, I stopped using it. Then another had the same problem in the same enclosure. How would I go about proving the enclosure is not the problem??

You're right about no eSata on the laptop. Which is a shame. To be honest, I've had a pretty good experience with prebuilt external drives, until today, and I suspect that was probably my fault rather than the drive.
 
But Old Hippie, I think the coolermaster enclosure was the problem in the first place
Could be.

Enclosures without active cooling (fan) tend to fry the drive inside.

Another problem could be the external's USB chipset which can happen to any.

I've never bothered to explore CrystalDisks details.

If I were you I'd be looking at Dell's refurb page :)

To be honest, I've had a pretty good experience with prebuilt external drives,
The problem is not being able to remove the drive for troubleshooting without voiding the warranty.
 
Thanks.

I suspect that is the reason for the problem. Temporarily high temperatures causing corruption of certain files. However, I'm now not sure if I keep the HDD at normal operating temperatures, whether the drive is ok.

I had a look at the crystaldiskinfo stuff and it seems like the drive seems to be alright. I suppose that is a limitation of SMART as it hasn't picked up all of these errors.

At the moment, I'm going to have to trawl through my existing backup disks to see if I can create md5 hashes to compare with my possibly corrupt HDD. This is going to be a real pain. If anyone knows of an automated process to check file integrity, please let me know. I don't know of anything other than just opening every file.
 
Also, if I discover a corrupt file that I want to recover from my HDD, does anyone have any suggestions for a good file recovery program?
 
I had a X-Craft 360 enclosure (the same one ?) with usb and e-sata connection and it's probably one of the worst enclosure regarding temp. IIRC it was discontinued quicky by coolermaster never to be replaced. According to the smart attribute, I had one caviar blue which ran hot with a peak at 95°C :D

According to the smart of your drives, the 1,5To got to 57°C and the 1To to 51°C which is hot but in specs (<60°C). Hard to say if it's due to the temp or not but obviously this is not the best for the drive.

All these enclosure without airflow must be used with a low rpm hdd or a 2,5" . I also would stay away from the pre-made one. Grab one with a fan and some nice connection like e-sata and or usb 3.
 
Thanks Wythix, that's good to know. I had a look at some of the enclosures with fans, but they are quite expensive as they usually take 2 HDDs. Do you know if I can put my HDDs in the Synology and then connect it to my laptop so I can run badblocks on it? Would then save my self the cost of buying a new enclosure.

95degC??? I am assuming your caviar blue is now dead? That's amazing if it is still alive and your data is intact. I am currently going through my files and a lot of them are corrupt. I would estimate around 10%. If my smart data has the temperatures within specs, then either this isn't the problem, or the smart thing isn't as robust. Or I didn't turn it on. Hmmmmmm, it's probably the last one, although I have no idea why I would do that and I have no recollection of turning it back on.

Luckily I backed up most of my important data, but there are lots of media files there that probably have errors. Unfortunately I don't know which ones. Anyone know how to check possible corrupt media files? Or is there a program that can just watch all of the videos and then tell me if there is an error in it?

I'm going to play with recuva now to see if I can pull out something reasonable from the HDD. :eek:
 
It's a disk from 2006 with both sata and molex connector which even with proper cooling was hot. The disk ran fine after that, but it was only used as punctual back-up and is now retired.

Seagate drives do have a threshold smart attribute with a max temperature at 55°C. A temperature above 50°C (?) should trigger a caution warning from CDI, but for what consequences ? There was a test from Anand with the first Seagate 3To in a closed enclosure. It ran excessively hot, and he mentioned possible errors on the drive http://www.anandtech.com/show/3858/the-worlds-first-3tb-hdd-seagate-goflex-desk-3tb-review/7

I don't have a synology so I let someone else answer about it, but you can also rip the enclosure apart and just take the small pcb and hook it up to the drive. There should be no more heat problem. Assuming that the controller is not the culprit.

For the comparison you can use rsync with --checksum or with a gui there's also freefilesync.
 
Last edited:
but they are quite expensive as they usually take 2 HDDs

I've made a external enclosure, active cooling list for ya on Newegg but it won't let me link the entire list. AAR how 'bout this one?

SMART didn't used to work with USB but it looks like they have that worked out but with some stipulations.
 
Thanks for all the ideas. Unfortunately, the only external enclosure models I can get my hands on are by a company called 'century' (Japanese) and something called Stardom??

My important files are backed up in multiple locations etc. The stupid thing I did was to amalgamate all of my media files onto this single HDD. I think I formatted the other HDD that used to contain half of them. :eek: I'm not sure how I can check that each video works other than by physically playing them and checking for errors. Ideally, they shouldn't even play, but that would be too lucky.

In the meantime, the single drive enclosures that are available in Asia don't have fans and the double drive enclosures are much more expensive. Something like 3-4x the cost of a single drive enclosure. I was planning to buy a Synology DS413 and wanted to know whether I could access the HDDs as if I was using an enclosure, without the Synology software interfering.

Thanks for the details with the USB. I suspect the high temps were not being recorded due to ?incompatibility with the USB controller. Lots of assumptions there of course. :D
 
If you look at the Synology DS413 spec page it lists JBOD and Basic as optional storage types.

I would think one of those would let you have separate disks.

I suggest you download the manual and check for sure.
 
Thanks! Great news. Time to empty my bank account. :) By the way, I saw there was a DS413j, but that only has USB2.0, which I'm not that keen on. What useful features for the home user does the 413 have that the j doesn't?

Also, does anyone know of a reliable CRC calculator?
 
Just checked a few files using the CRC calculator and they all seem to be alright. But I am a bit doubtful about this. I assume using MD5 is better, but is CRC alright if a file is only a little corrupted? I hope so, as I don't have anything else to check them with. :(
 
Should be a no-brainer to go to the Synology site for all your questions.
 
I had a look at the Synology site, which was quite helpful, and there were a few things I picked out, but most of the software details seem to be similar. Having said that, compared to the 413J, the 413 has hardware encryption and also USB3.0. Has anyone else decided between 413 vs 413J and which did you go for?

Any experts out there on CRC or checking a video runs correctly without watching the whole thing, would be grateful for your help. I do wonder whether I could try to re-encode a video via Handbrake or something and if there was a corrupted segment, then Handbrake should fail. Is this a reasonable thought?
 
You already have to have a valid CRC on the uncorrupt movie first to compare it to.

Unfortunately you have to watch the whole movie. If the file header is intact then any other part of the file can be corrupt and it will still start to play fine.
 
I have a valid CRC on some of the videos, but not all of them. I found out that I can calculate the current CRC value using teracopy. However, around 10-15% of my image files are corrupt as they look weird, so I am assuming that almost every video will be bad (due to size of videos and distribution of damage). But the CRC checks for the first 40 or so have been fine. I find this hard to believe. Since I am manually comparing the calculated CRC value to the one I have stored, this is taking me a long time and I would rather not do it if I am doing something wrong or have assumed something incorrectly, which is probably the case :confused:

I also found this thing called divxrepair that supposedly goes through the avi file and tells you if there are any segments that are corrupted, but it didn't display what I expected it to. Very odd.

Anyway, from the sounds of things, I'm up the creek and will have to set myself down for a long long video session :D
 
It is nearly impossible for random data corruptions to screw up a movie file while still producing the same CRC checksum, it is not (much) less safe than MD5. MD5 is a cryptographic hash, which means it is safer against someone who is actively changing data while trying to keep the same hash, but against random changes CRC32 is safe enough.

Video container formats are usually engineered in a way that allows some corruptions in the data stream and still be playable, although with frame errors. A proper player or reencoder should be able to resync to the stream if the file is not completely screwed up.
 
I have a valid CRC on some of the videos, but not all of them. I found out that I can calculate the current CRC value using teracopy. However, around 10-15% of my image files are corrupt as they look weird, so I am assuming that almost every video will be bad (due to size of videos and distribution of damage). But the CRC checks for the first 40 or so have been fine. I find this hard to believe.
One possible explanation could be that the files were already corrupted before you did the CRC checksum?
 
Thanks for the tips. It's great news regarding the CRC check. They should be good as I watched them at the time the CRC was documented, so they should be good. Still it's kinda weird, but it's a risk I'm willing to take.

This whole backup thing is giving me a headache. I have different versions floating around in different hard drives. As bigger HDDs come out, I buy it and include some of the old stuff and a bit more of the new. I'll see if I can buy the Synology tomorrow so I can do a bit of re-organisation.
 
Ok, just an update for the first of my WD black drives. Ran it through 1.5 passes of the destructive run of badblocks, which looked like it had some errors. Crystalmark showed a current pending sector of 1, which seems to have produced a tag of 'caution'. Then I used the WD lifeguard, which also found some surface errors, but when I tried to repair it said there was an error and it couldn't fix it.

So this is my first RMA and it is a bit of a pain. I bought it in Hong Kong and it says the drive is out of region. Even worse, the serial number and PCBA don't seem to match. It was a few years ago now, so I think the receipt has gone and I can't remember which shop I bought it from. May need to give WD a call to sort this out, or else just bin it.

So my question is, do I need to run a secure drive eraser on the disk? I'd prefer to use a windows program so I can carry on and do other things, so would CCleaner do as I saw it has this function. Or should I just keep with partedmagic? Thanks again for all the advice.
 
Just to let you all know. Managed to get a RMA on my 1TB WD black. Seems to be quite straightforward over here in HK. Just popped over to the agent (Synnex), showed them the 'failed' drive and evidence (lifeguard image), then they swapped it with a new drive. Well, they say it is brand new and crystaldisk didn't show any reallocated sectors. Just running through badblocks now to make sure all is a-ok.

My other 1.5TB drive managed to stay ok during 2 passes of badblocks and the extended lifeguard test, so it is currently in use. Fingers crossed it lasts many years.

Will probably buy a synology NAS eventually, but not just yet.

Thanks again for all the help and advice. Much appreciated!!
 
Back
Top