Why You Should NOT Use RAID 5 Storage ( But Use RAID 6! ) 

A lot of people are having trouble understanding that yes, a single URE can indeed cause the entire array to fail to rebuild. Because of the nature of the parity system, it is NOT the same as having a URE on a standalone drive, where only that single file is lost. You need it ALL to rebuild it all. You might be able to send the drives to a data recovery service who could potentially rebuild the array and only lose the files directly using the lost parity data, but that would be very expensive and time consuming, and not necessarily a guarantee. Not to mention that because array rebuilds take so long, and are very very intensive, it drastically increases the chances of having yet another URE in that time, which could completely ruin any chances of any recovery at all. You can google something like "Raid 5 rebuild failure probability calculator" if you don't believe me on the odds. But using 10^14 URE rate drives is extremely risky and you actually are VERY likely to experience a rebuild failure even with small arrays. In any case, when looking to buy drives for a NAS, my suggestion is ONLY buy a drive with a URE rate equal to or better than 1 per 10^15 bits. This can be found by googling for the 'Data Sheet' of the specific model drive you're looking for. It will be called something like "Non-recoverable read errors rate", "Error Rate (non-recoverable)", "Unrecoverable Bit Error Rate (UBER)", or something similar, probably listed along with the 'reliability' specs in the data sheet. I haven't really been able to even find any HDDs with a URE rate better than 10^15. However, with SSDs you can usually find better URE rates. Data sheets for consumer desktop SSDs (Like samsung's EVO/PRO SSD line or Seagate's Barracuda SSD line) don't list URE rates. But Seagate's "Ironwolf" NAS SSD line actually have URE rates of 1 in 10^17, so 100 times better than any HDD I've seen. However those cost several times more than an HDD (right now it's $750 for just a 3.8TB drive).

I would argue that the bigger lesson here isn't to get overly hung up over RAID5 vs RAID6, but to remember not to fully trust either of them and always have a proper backup. A good backup system turns a nightmare into an inconvenience.

For small ammount of disks (like 4 or 6) usefull space / used space ratio is too small to consider RAID 6, you just need to stick with WD Red, Seagate (not Barracudas!) Ironwolf / Constellation and similar kind of disks and you'll be fine. For larger ammounts, you use ZFS anyway to avoid bit rot. The only really dangerous scenarios i see are some instant power failures and thus you need either RAID controller with battery or UPS.

People have different storage needs. There isn't a one size fits all solution, and an absolute statement like "RAID 5 is bad, never use it" is just silly even today. RAID 6 _is_ safer for large arrays where more disks means more points of failure, but for smaller arrays RAID 5 is still fine. It may even be preferable because it has better storage efficiency. Besides, you should be backing up data that's important anyway. RAID is only meant for minimizing down time when a drive dies. If you're ever using it by itself as a substitute for a true backup then you're an idiot.

it doesn't matter if the raid fails because you should have a back up of the raid itself.

Thank you for this vid. Now I backup my data 'cause of bad sector, tried raid 5 twice and it always have error, will try this raid setup :D

I lost a primary domain controller once back in 2003. The raid 5 didn't help, neither did the secondary domain controller. It was a huge mess. I had to accept defeat and rebuild everything from scratch in two weeks with the help of a friend. Slept at the office for that time. So now I am a bit skeptical to all redundancy systems :)

I've been using a 12 drive Linux Software RAID5 setup the past 10 years. Gradually growing and replacing drives (from 8x200GB to my current 12x2TB). Although I agree with you that RAID6 would be safer, I've never experienced the problems you're describing. I've had 1 drive failing periodically, and replaced maybe 4-5 drives in total (due to failed/dead drives) during the same time period. This is data that I am comfortable loosing, although I would prefer not to. The worst problem I've encountered during this time, is the failing of a controller, where symptoms indicate that multiple drives are lost. But thanks to Software RAID, I'm not bound to a specific hardware disk controller. Which makes it easy to rebuild or even move the entire array to different hardware. So if I encounter the bad sector issue I'll report back when that happens, although I haven't seen a bad sector since the mid 90s

For the record, raid 5 is still fine for small arrays like an executive workstation for your CEO. His premise here is that you're using a very large array like in a NAS, Backup array, or Data Center

I've rebuilt a few dozen 20~30TB RAID 6 arrays without a single rebuild failure. RAID 6 is solid. This has always been with Adaptec controllers (51645, 6805, 71605 etc.)

Normally if you create a raid set you also set a spare drive so it can start rebuilding the raid asap. Depending the raid you want/need it will be better one or the other. Also keeping in mind performance of course

I have read that for a raid with multiple hard drives next to each other, you should only use NAS or Enterprise hard drives as they have vibration detection. Because cheap hard disks without vibration detection are actually only intended to be alone in the PC case or individually in an external case.

But how do I download more raids?

There is something I don't understand about this: Assume you were to use 12TB HDDs with one URE at every 10^14 bits, your probability of an URE would approach 100% during a RAID-5 rebuild. However it would also approach 100% during a RAID-6 rebuild, because every HDD itself would likely have at least one URE. So no matter how many parity HDDs you use, you are always hosed and thus RAID-6 would give you anything over RAID-5. What am I missing here?

I understand RAID 6 has a slower write speed than RAID 5. What about RAID 5 plus a hot-swap drive? That's what I'm doing right now. So if a drive goes bad the hot-swap kicks in and it rebuilding the array automatically. I replace the bad drive with a new one and it becomes my new hot-swap. You get part of the benefit of RAID 6 (although not the exact same) without the performance hits of 6 vs 5.

I have an Asustor 2404TE, one of the very early line of Ausstor NAS that had an early Atom CPU. It had a 4x3TB setup (WD Reds) at RAID-5. The 3pin cable conencted to it wore and there was a power surge that caused damage to one of my drives. Eventually it was reporting a lot of rear errors on the drive, which started small then began to build up. I don't know if they were UREs, but there were many. The system rebuilt the raid twice with no issues during this time. I then replaced the faulty drive (4) with another, the RAID rebuilt. That was back in 2017 and there's been no issues since, though the other drives will need replacing soon. So while there's a risk of URE it's incredibly low, and your drive will likely start reporting errors before this stage, at which point you should replace the faulty drive ASAP. Raid 6 has more redundancy, but I don't think it makes up for the exorbitant cost to get the same level of storage. I am planning to get a new ASUSTOR with a 4x4TB setup. Obviously my rebuild risk is higher, but I don't think it's worth me forking out another $400ish just for the same level of storage.

Great video. I'll be using RAID 6 on my new storage! Many thanks for your effort, cheers!

ZFS is the future of drive arrays.

A lot of newer harddrives like the ironwolf pro have one URE in 10^15 bits instead of 10^14. So that means 125 tb of storage where one bit will fail you.

I'm not sure this advice is particularly helpful for a SOHO user. His implication is that that your RAID array somehow forms part of your backup strategy (or perhaps even *is* your backup strategy) which in itself is a poor message to convey. I really can't see too much issue with RAID 5 + proper backups. How many SOHO environments need 2 levels of redundancy unless in special situations?

Nice video, As a systems engineer Raid 5 and or Raid 50 is discouraged partly because of what you mentioned in your video, but also what a drive fails in Raid 5 and or 50, the drives spin up to full read speed when the bad drive is replaced. This full read speed increases the chances of the other drives failing during the rebuild process and until the new drive is completely rebuilt your chances of failure are much higher. The errors that will leave a drive useless are much more that what was mentioned in your video.

Ok so I can use RAID 5. Thanks

personaly i know about a risk of loosing data from RAID5 during rebuild cuz drives are 100% loaded ( work+raid rebuilding) and have bigger chance to fail at this time, so basicly you can loose data on RAID5 during rebuild process, hard to tell if its cuz of URE or cuz something else, but RAID5 is ok basicly to 4-5 drives max, everything above is for RAID6 or if you can afford it, i belive none on common users cant( and dont even need it :D) then something like RAID60 or w/e

I used raid 0 to install my steam and origin games on, since saves aren't usually stored in the install directory, and I have no data cap. it wouldn't be a big issue if the raid failed, and it's a lot cheaper than buying a SSD big enough

another thing to take note is when a RAID 5 is rebuilding it works your existing(possibley near EOL) drives harder to re build the array. I have had 2 drives fail on me in the middle of a rebuild

I think for a bunch of mystery-meat disks you found lying around in the storage closet, yeah, use RAID 6. But if you're not fucking around and actually trying to run a business: use one big RAID 10 combined with a real backup regimen...

Is all the data really lost though, with the read error? Wouldn't you have like one or two corrupt files, but still get the rest back?

I was setting up raids before you were born but you were clear decisive & I actually learned something about RAID 6 vs 5 from you nice brother !

excellent content from the Michael Cera of Tech :D

Unrecoverable Read Errors (UREs) are rare unless you're using Seagate drives. I'll very buy Seagate drives again. UREs is what killed my last RAID 5 array. One disk went bad but I could not rebuild because the other Seagate disks all had UREs on them.

well you can just use RAID 10, simple, efective and reliable.. i know you have less space.... but its more secure for critical data ;)

This guy is right. A couple of years ago I nearly lost data because for whatever reason one drive died and the other had lots of errors. ZFS told me exactly which files were unrecoverable and thank god I had a backup for them. The backup was old so if other files were affected they might not have been in the backup and therefore then been lost. After fixing the RAIDZ I migrated to RAIDZ2 which is equivalent to RAID6. Additionally I copy the really important files daily to a backupdrive using snapshots and replication. Additionally i advise only to use drives that are made for NAS. They feature TLER (time limited error recovery) which is very important in a RAID in error situations. When I had the error mentioned above I didn't have these drives because they were only available for enterprises and extremely expensive. With upcoming of home-NAS to the mass market this changed. So avoid desktop drives for NAS.

That is why we use Raid 5 and 6, but need to have a Backup. I see your point. But I cannot see your conclusion. RAID and backup are not the same thing

Hey Joe, lets us consider the following scenario. what is I have 10x8TB array of disks configured with RAID6. So even if two disks fails( Since RAID6 ). I will be having 64TB of data to be scanned for recovering the last drives. So from your URE probability calculation, there is URE happening. Then RAID6 is also not useful. What would you suggest about this... !!!! Thanks in advance

i can no longer take this guy seriously i just assume its bullish

You are aware that you lose performance in a raid 6 array. I work for an MSP that specializes in dental I.T. These places due to Canadian standards must keep data for patients for 10 years. Every one of our 2000 servers in play all have a raid 5 arrry and for the most part there are no issues. Now keep on mind even with a raid array we still run a cloud and on site back up of the host and the DC VM and Server VM. You should never %100 rely on you raid array.

Thanks, lot's of people say don't use RAID5, you're the first to explain why!

trust issues.

Great information. Very useful video 👍. In the future, I will look for (quoting you): "Non-recoverable read errors rate", "Error Rate (non-recoverable)", "Unrecoverable Bit Error Rate (UBER)", or something similar, probably listed along with the 'reliability' specs in the data sheet. Rated at 1e15

Is RAID 10 faster than RAID 5 and 6?

That's why I use Ultracopier with "verify checksums" activated. I use a small NAS with 2 8TB drives in RAID 1.

With RAID 5 and a hot spare don't you get the same result as RAID 6 except you have less probability for URE because you have less drives in the RAID?

Wait wait, is this his joke channel or his real one lol?

and its compounded in that ALL of the drives have EVERY sector read during a rebuild for maximum heat and least reliability. With 8 * 18TB your talking about 24 hours non-stop to rebuild

I know this was a few years ago, but correct me if I'm wrong. This really depends on the number of drives that you have? Right? What if I'm only using 4 drives instead of 7? This will certainly change my chances a bit. I agree with what's said, but the scenario certainly depends on a few factors.

Just use 6 drives in raid 0 for maximum craziness :) :) :), I live dangerously...

One should be backing up their data regardless of what RAID configuration you are using. Yes, RAID 5 is riskier than RAID 6. That risk may or may not be acceptable depending on your situation though.

I've read all over that read speed is fast on RAID5 and RAID6, but due to the need to write parity data across all drives, the write speed is very slow, even with certain hardware RAID Controllers

His theory has a slight point. His problem is that the entirety of his experience seems to just be theory. In practice, RAID5 works just fine and people have been using it for a very long time without issue. I've been doing it for around 20 years and have yet to lose any data I wasn't able to recover easily. His descriptions of RAID levels is woefully inadequate. RAID0 (striping) is used for speed. The data can be lost as it is normally backed up regularly (usually to a RAID5 array). RAID1 (mirroring) is used normally for either cases where you are only using 2 drives for storage and want redundancy, or (as I normally do) to mirror the OS drives on a server. RAID5 and 6 are used for storage arrays. He also fails to mention RAID10, but that's really just for the big enterprise users. In reality, the typical cases where you use RAID6 is in a situation where you have an array is a location that is not attended. A remote datacenter is a good example. If you lose a drive, you don't have to get up in the middle of the night to go swap a drive because you have that extra peace of mind knowing you can lose another and be fine. RAID5 is used more when the array has someone nearby that can swap a drive out rapidly. The exception is if you have a server that has, for example, 25 drive bays (like the HP Proliant DL380P). If you have 25 bays, you can more afford to lose the capacity. Most SMBs cannot field those servers due to cost. They're mainly for enterprise users. If you have the typical storage server for SMBs that has 8 bays (2 mirrored for the OS, and 6 for storage), it's harder to justify that capacity loss. If you have a device with 4 bays, RAID6 isn't even an option, while RAID5 is. Another thing to think about is that you want to have offsite backup for any storage array. Having this backup will also insure that if your array happens to fail, you can restore your data from that. In the end, it's preference. If you are more the paranoid type (not a bad thing), or have a large budget for more storage devices, then go ahead and use RAID6. Otherwise RAID5 is a tried and true method of housing data.

Hi, that's right in theory. Regarding your reason, it would also make no sense to build RAID 0, RAID 1 oder RAID 10. Because all of it would not be safe enough.... at least in theory. But in practice, good array controllers do in small intervals surface scans and analysis so you don't have to worry about URE. And by the way, RAID 1 does typically not increase read performance! Cheers

Do not use RAID for storage! If you are using raid 6 you are "walking on thin ice" as you say. I'm not saying you didn't do your homework, Joe, everything you said was right. However, RAID by itself is not a good idea for a storage system, and you actually touch on the right idea why: read errors and bit rot. Take this as an example: When you have your RAID 6 set up and go to look at an image you stored last Christmas, what is to say that your storage array gives back exactly what you put in? Fine the disks can have some degree of error correction but what happens when, over time, some bits flip because of a cosmic ray hit or something like that? Well, you get the wrong data back. Mostly it would be a pixel block in a jpeg that turns a shade darker or something unnoticeable like that, because that is where the probability lies. But once in a while, something worse could happen causing the loss of that image, or video for instance. So what's the solution then? Well, it is the file system that you employ. BTRFS and ZFS are two of these on the linux platform and ReFS by microsoft for servers is a relatively new one. The important part here is that these file system supports something that is called checksums which helps the file system figure out if something is broken. And without going into too much detail, makes the file system able to get the correct data from a pairity block (yes a raid but inside the file system). Go check out TekEnterprise's video "RAID: Obsolete?" for more info on that... So what does this mean for the average user or slightly data hungry users like photographers and people who like to download a lot of things from the internet all the time...? It means don't. Get a single big drive, and designate it as your backup drive. Then pay for a service like Amazon Glacier or Backblaze (i like backblaze's business model). What raid is good for by itself are things like cashing and fast access memory that is none-vital. It is NOT AT ALL a backup solution. Gamers (and even photographers) who want quicker loading times can use SSD's in RAID (as long as vital data is backed up). And if they need more of that ultra fast storage they can get 4 ssd's and put them in RAID5 so that their downtime would potentially be less if a drive fails. damnit i wrote too much again...

I use BTRFS on top of my RAID 5, so an invalid bit can get detected. When there is a read error, it will be just a single bit flipped out of some terabytes. And as I said, that bit will not be a big problem since BTRFS uses cluster checksums and can repair an invalid bit.

Do you ever blink?

Thanks, good information. I guess my only question is: How do you archive this much data when you do not need it on a local drive? Such as old youtube videos. I want to keep them, but not locally when I use up my 20-30TB NAS.

Everything I have read says that even RAID 6 is asking for trouble - if you care about your files, you should use RAID 10.

And if I use a RAID 50? Two RAIDs 5 in RAID 0?? Such a waste of drive ah?? And if I use 2 SSD in RAID 0 as virtual memory?? It could work enough fast as a RAM Memory?? Very good explanation Joe, greetings from Argentina

your out to lunch!!! you don't know what your talking about!

That's only true if the overall system is so big that you can't afford to have a backup of everything. Usually if you care about the data and it's very important you can't rely on any raid to save your data. If it's the same physical location you have a lot of risks like - fire, power surge, user error (delete a file by mistake), software error that corrupts your data and others. That's why you should always have backups and they should be incremental (with history). So if you have complete backup (to other system, cloud or whatever in different physical location) then not being able to rebuild raid 5 is not big (that) deal. Of course if you have 6 or more drives you would like to have raid 6 or two raid 5 to save time in case of fail+read error two drives failure. But raid 6 on 4 drives would be overkill.

now i had a full idea why some of my old data become unreadable, and in the first place i did not see the deference between RAID 5 & 6 and you make it very clear thank you very much RAID 6 is the best in my opinion until now, and i will keep looking. after all that can the user use same two drive specification with two different brands or there will be a problem with read write performance

Or... You could also switch over to a flavour of BSD, with native support for zfs, and thereby enable parity-driven raid, with as many parity drives as you want. Or of course, just use the many other possible flavours of raid or "raid-like" configurations. You also have the advantage of not being hardware-locked into the same raid-adapter (model and/or make), if you find yourself in the need of getting a new one, if the old breaks down.

Maybe use Raid 5, along with the batteries Iv'e taped to my cat-5 cable, and the water bottles on the antennas. I should be able to get my modem to launch, and maybe I'll make a 3 TB Netgear drone quadcopter. Put like, a sick ass skull paint job on it.. Lol...

I just had a disk fail in a 4-drive RAID5. Swapped the drive and it automatically rebuilt it - no problem at all. Been using RAID5 for many years and can recommend it. RAID6 would only make sense with 5drives+ dude.

I wanto create a RAID 6 array with 5 disks in Windows 10, but HOW do I do that? As I understand it windows 10 doesn't do RAID 6, and so far I haven't found any reasonably user friendly software that does.....PLEASE HELP :-)

If you are using a decent RAID controller / software a read error will NOT cause the raid rebuild to completely fail and mean that the whole array will become unreadable as you stated at ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-A2OxG2UjiV4.html A decent RAID controller (such as any PERC one - I have had a raid rebuild come across read errors while rebuilding without serious issues) will mark blocks as being bad when they are found to contain read errors, and continue building the rest of the array, and log the errors. What does this mean... the same thing as if you have a read error on a standard HDD partition, the partition effectively gets a little smaller, as the sector is marked as bad and avoided for reads/writes in the future... any files that were in that sector will usually be moved to another sector with those parts in the bad sector filled with the ASCII(0) character.

My raid is my backup. The odds of losing data on my primary computer or the entire drive and a drive on the raid and having a ure at the same time is low enough not to worry about.

Isn't raid 10 better in that scenario since you get the redundancy of raid 1 but with the speed of raid 0 as well as being able to loose 2 drives and still be able to recover? Isn't that supposed to be the best of the three worlds?

Joe very good and informative vid. I am a Network Administrator and I have a question for you that I would like for you to answer. I already know the answer to my question BUT their may be viewers that may have the same burning question. Ready? Here goes - Joe I have a WD NAS, or some third party NAS or a PCIe Raid card that does not have a RAID 6 option. I see RAIDS 0 - 1 - 5 - 10. Which do I use since 6 is not available but I want and can afford the drive cost of RAID 6 what does one choose Joe? And Why? Thank you again for the very good vid.

Raid 6 on 4 Drives cuts drives in half like Raid 1, right? so Raid 5 then… if Drives are more then four (>4) then Raid 6... did I understand that correct?

Just subbed and Thank you for your informative videos.

I understand the "You can not rebuild" but if i run Raid 5 and 1 drive fails, i will still have access to the data even though the raid is degraded. Would it not be possible to do a full backup of the NAS before trying the rebuild. ?Also i run Raid 5, but i also have another NAS with a single disc, where i have an Automatic backup job set up so i have the data on that disc as well.Also the Raid 5 NAS is in my locked shed, and the other disc is inside my house.This way it is really unlikely that both Buildings would burn Down (Shed is not next to the house) and also it is unlikely that both NAS'es would be stolen. The thief would set off the alarm when entering any of the 2 Buildings and would not have time to rob both.

Post 1987: RAID = Redundant Array of Independent Disk Pre 1987: RAID = Redundant Array of Inexpensive Disk

I bought 4 drives 14 TB WD Ultrastart and QNAP what raid can I apply to have more space and not experience UTE .Thank you

thanks a lot mate. I did not know what the difference was and what parity was. :) kudos to the video. very nice.

RAID has many more problems like bit rot and power fails during writing. Any serious commercial centre will use ZFS for storage.

can raid 5 be converted to raid 6 in windows 10 without losing data?

windows server 2019 dosent support raid 6 how can you set it up thanks

If you're just building a NAS for movies and Steam Games then RAID 5 is fine and if you lose your data then so what just download it again! If you have mission critical data that needs to be protected then an offsite back up plan using multiple back-up devices and or a cloud based solution with continuous syncing of data is what you really need!!!

The problem with raid is it doesn't protect against user error or file corruption it's not really a viable backup solution.

@Thios.Joe Tech. I just picked up a 4 dri.. It just seems so wasteful because RAID 5 practically speaking gives you 2 drive redundancy but you lose a lot of storage, I don't need more than about 8 T. I thought to use 4 4 T drives which would give me 12 T of storage in RAID 5....but only 8 with RAID 6.....seems like wasting a lot of storage. I have also heard that one should routinely scrub the drives i a NAS....not sure if that is more than SMART or some other technology.

for home use Raid 5 is fine. specially because it can be done with 3 instead of RAID6 4 drives. Just set up a weekly/monthly error and repair schedule.

if a 2TB drive died, surely you would only need to read 2TB of data to restore what was on the failed drive. in your example that would still give a 1 in 6 chance of a URE though, which is still too high for my liking edit: drives like the WD Gold have a URE rating of < 1 in 10^15 which is less than 1 URE every 125TB. In your example even if all 12TB needed to be read the chance of failure is

Raid 60 is not mirrored it's stripped. It's two raid 6's put in a raid 0.

Great Explanation of RAID. Clear, Concise, Interesting.

What is your opinion using SHR-1 with BTRFS with four (4) 6TB NAS (Iron Wolf) disk drives in a Synology DS918+?

Joe. Thanks for the Vid. Question: If you get an URE on a 4 disk RAID 5 during rebuild, is that disk array still readable, assuming a second disk hasn't crashed? i.e can you still copy off the data onto a backup?

How does raid 5 or 6 increase write speed?

Raid 5 works fine, my only concern is the rebuild time of the larger hdd's I use raid 1+0 never had any issues

Thanks! Just in time as well as im building my own raid with my hard drives.

I use RAID 0. I have no need for RAID redundancy. RAID is no backup anyway and if a drive fails, I have backups to restore from so then I can just replace the failed drive, rebuild the pool and restore from backup. RAID redundancy for me just seems unnecessary and cumbersome. It means less write performance with parity and it isn’t worth anything as rebuilding with parity takes time and you can lose more drives during the rebuilding process. The best RAID that I see for fast rebuild is RAID 10 but even with that, you can lose both drives in the same set so then you’re screwed anyway. It is also not cost effective as 50% of the capacity is lost to mirroring. I rely on backups. If the system goes down, I just rebuild the whole thing from backup. RAID redundancy is solely for people who need to get their system up and running fast without having to go to their backups. If you don’t need that, RAID redundancy is just an unnecessary extra thing to fret over. Make it simple and go with RAID 0. You get the performance and you don’t have to stress over what RAID you should and should not use. You should always have backups either way and those are generally enough without going the extra mile and have RAID redundancy. It’s just extra complication as I see it.

So.....your saying I shouldn't put my 8 4tb ssd's in raid 50?

OK, so assuming that the calculation of a 71% success rate to rebuild a RAID 5, 4x16TB drive array, then RAID 5 is a no go. Here is my question then. Which on is better: 1) ONE SET of 4x16tb drives as RAID 6 OR 2) TWO SETS of 2x16TB drives as RAID 1 I would almost guess #2 right ? Reason being that one day, if a drive takes a total dump, right out of the gate, only 1/2 of your data is at risk (1/2 of 32TB). Even the at risk 16TB of DATA in the compromised RAID, the chances of BOTH DRIVES failing COMPLETELY is gotta be close to nothing. And even if the 1 remaining disk that you hoped is good, if that one turns out to have more than few UREs on the drive, not all the data is lost. You should still be able to use a low level data recovery utility to get a majority of that data off that drive at the very least, correct ? So, really if the worst of the worst of the worst happened, you should have confidence that say 90% to 95% of all your data will still be available. Is there something that I missing here, or is there another big advantage to going with option 1 - RAID 6 option ? I appreciate the input and I appreciate your video !!! Can not get over that high probability of failure ... that's just crazy !

To normal users though a regular backup and using decent hardware is the cheapest option. Set it to run overnight...If you don't want those hours of gaming progress lost then save to 2 cheap locations and forget raid. Large companies have raid systems in place but also have geo redundancy - ie fallback from raid is a fail of all HDD/SSD's connected = total loss. Does raid 6 put a fire out? no so this is old thinking but stil used.

Would S.M.A.R.T give you any warning of impending URE?

+ThioJoeTech Thanks for the informative video :) I'm confused with RAID 1. I want mirroring and also better read speed (write speed is not that important). Is RAID 1 always guaranteed to have double read speed? How do we configure RAID 1 to guarantee double read speed?

Now I feel guilty for running Raid 5. Wishing I did 6.

two main issues with this video: 1) there only very few controllers supporting raid6 in the consumer range - most controllers only support raid 0, 1 and 5 and combinations of them (although often only 1+0 / 0+1) 2) as windows still has the monopoly it only supports striping for mirroring on client versions - and even windows server only support raid 5 out of the box - so, if you want to build raid6 with windows you have to rely on 3rd party stuff unlike linux - so, sure, raid6 should always be used over raid5 - but it's often not possible to do other than soft-raid on linux ...

got the same problem... how can i block him

Hello ThioJoeTech, Thanks for making this video. I'm having a hard time finding URE: Unrecoverable Read Error in specifications for hard drives on websites such as Newegg and Amazon. Any way you can clarify how to find this information? Also, do you know how reliable Windows 10 is for setting up a RAID 6? To start out, I just want to have 3 12 TB drives in a NAS Enclosure. I'm looking at the Seagate IronWolfPro 7200 RPM drives. And I'll just back up to a BackBlaze account. I'm currently using 3 Google Drive accounts to store my archived projects and it's getting costly and a pain to upload. Sorry if this comment is long, it's a bit difficult finding the correct information and way to go about this. It is an investment, and I want to make sure I do it the best way possible. If you could let me know I would appreciate it. Thank you. - Paul Harwood

Sir, i f i may ask, if you copy or transfer files while Raid 5 is rebuilding, will it affect the system?

Great video as always. Watched your video on the Sony FS5 and it was just what I needed to know to dump my Canon 1 DX and buy the Sony