They sent my flawed product back.
ฝัง
- เผยแพร่เมื่อ 3 พ.ค. 2021
- Check out the DROP ENTR Mechanical Keyboard at dro.ps/ltt-may-entr
SmartDeploy: Claim your free IT software (worth $882!) at lmg.gg/yMhY8
My RMA return from Intel finally arrived so I'm going to find out if they solved the problem I was having with these storage drives and if so what they changed.
Buy Intel P4500 SSD (PAID LINK): geni.us/1ZRtl
Purchases made through some store links may provide some compensation to Linus Media Group.
Discuss on the forum: linustechtips.com/topic/13344...
►GET MERCH: www.LTTStore.com/
►SUPPORT US ON FLOATPLANE: www.floatplane.com/
►LTX EXPO: www.ltxexpo.com/
AFFILIATES & REFERRALS
---------------------------------------------------
►Affiliates, Sponsors & Referrals: lmg.gg/sponsors
►Check out our podcast gear: kit.co/linustechtips/lmg-podc...
►Private Internet Access VPN: lmg.gg/pialinus2
►Our Official Charging Partner Anker: lmg.gg/AnkerLTT
►MK Keyboards: lmg.gg/LyLtl
►Nerd or Die Stream Overlays: lmg.gg/avLlO
►Official Game Store: www.nexus.gg/ltt
►Amazon Prime: lmg.gg/8KV1v
►Audible Free Trial: lmg.gg/8242J
►Our Gear on Amazon: geni.us/OhmF
FOLLOW US ELSEWHERE
---------------------------------------------------
Twitter: / linustech
Facebook: / linustech
Instagram: / linustech
Twitch: / linustech
FOLLOW OUR OTHER CHANNELS
---------------------------------------------------
Mac Address: lmg.gg/macaddress
Techquickie: lmg.gg/techquickieyt
TechLinked: lmg.gg/techlinkedyt
ShortCircuit: lmg.gg/shortcircuityt
LMG Clips: lmg.gg/lmgclipsyt
Channel Super Fun: lmg.gg/channelsuperfunyt
Carpool Critics: lmg.gg/carpoolcriticsyt
MUSIC CREDIT
---------------------------------------------------
Title: Laszlo - Supernova
Video Link: • [Electro] - Laszlo - S...
iTunes Download Link: itunes.apple.com/us/album/sup...
Artist Link: / laszlomusic
Outro Screen Music Credit: Approaching Nirvana - Sugar High / approachingnirvana
Monitor And Keyboard by vadimmihalkevich / CC BY 4.0 geni.us/PgGWp
Mechanical RGB Keyboard by BigBrotherECE / CC BY 4.0 geni.us/mj6pHk4
Mouse Gamer free Model By Oscar Creativo / CC BY 4.0 geni.us/Ps3XfE - วิทยาศาสตร์และเทคโนโลยี
maybe the dell warranty you got from secret shopper will help
2 things I don’t understand in life
1: Mathematics
2: *Why people don’t give me a chance to make it on TH-cam even though I work so hard making good content*
Plz help
@@eugenesmemes because you do shit like this
Because I'm not famous like other singers that's why no one see my singing videos. Just see once. ❤️.. ..
@@eugenesmemes Literally none cares fricknite bot…
@@eugenesmemes because you play fortnight. One of many problems
It's so funny how Wendell is the Tech Support for a Tech Channel - love that man
That's because Wendel is a real DevOps / SRE
@@Level1Techs wendel don't you have better things to do other than commenting on TH-cam :p love your work
Ever hear “tech support for the tech support?” Same energy.
Fun fact, Ryan is the real brains and he lets Wendell take all the credit.
"you guys want a free tech tip?"
Me: "I've been waiting years for this"
Budget mem
yes
Third
Victoria's Secret? No I'd rather know Linus' forsaken Tech Tip
"Assuming Intel's Serial Numbers are sequential"
I mean, if a Serial Number is non sequential, is it still a Serial Number? Might be a Parallel Number instead lol.
Well, it's been common for military equipment to have non sequential serial numbers to obfuscate the number of devices issued if some of them are captured, thus also obfuscating the number of soldiers deployed in the theatre. I wouldn't say Intel is quite that hostile to its customer base, although it's getting there.
Or maybe a prime number?
@@MrVeps1 Ah, yup - the intel tank problem. Only becomes needed when thinks get really sour between them and AMD.
Non-sequential serial numbers have been a thing since the beginning of serial numbers. It's the numbers of the things in the series. Not the numbers of the things in the sequence.
Most serial number contains information like the production date (sometimes year, month, day but some other times a number of days or weeks elapsed after they started the production); other things that can be added are the batch number, the number of the production line, the number of the production plant, the hardware revision, the firmware revision, the capacity and many others things.
This is very common for many things.... tbh, these codes contains quite a lot of useful informations and are way better than a simple progressive number.
“What did they change?”
The serial numbers
I'm super serial in this one you guys.
They send another product in case of defect if they didn't want that hard drive shouldn't have bought it
@@just_hallen2999 so if you buy something and it turns out to be incompatible with the things you already own, even if they should be compatible and you couldn't have known otherwise you won't complain??
@@JL-pc2eh You don't have to complain. You just return the incompatible ones back to the vendor, and get something different. Unless your vendor is just some random guy on eBay. But even eBay vendors sometimes have good return policies. If it was something he bought from me on eBay, I'd accept the return, and send out a replacement, with the appropriate financial considerations for the different in cost. I'd just have to relist them as gently used, or whatever colorful language to say that they were used once.
Gigabyte: Oh Sh*t, he still has that, good thing he uploaded this random video so that we can get it back
14:30 that was a good one
It's such a fast server haha I wish they could do something with it
Would be nice if that shit would work properly. It is funny how Intel got the roast, when they did nothing wrong
@@Rainaman- I mean if it is only with one line of ssds and they are from intel wouldn't that by default make it a thing with the ssd? I personally would have loved to see him proof that other ssds work fine cause unless I missed it he hasn't proven the statement that it is only with this ssd.
@@bb2ridder757 Except they're second hand drives not purchased from Intel, being run on an engineering sample. Who knows if the release version of the board even existed at the time the SSD's were in production. The issue is the use case Linus is trying to deploy them into. Now if Intel had certified them to work in such edge case arrangements, then that would be a different matter.
Congrats to the audio editor getting Linus through the "it's whisper quiet" server shots
@@elitepro3176
Stop it.
Go and scam somewhere else.
"They are different drives"
WD: * slowly moving new drive stickers with refurbished text and new date code behind their back *
This sort of subtle incompatibility is what drives people crazy working in IT. Diagnosing stuff like this is a nightmare.
If you use servers from major vendors and parts approved for them you don't get issues like this. If you cheap out and try to run off-brand engineering sample server with second hand SSDs from eBay in production, that's the price you pay for it.
I actually work at Intel and although I haven't seen it myself I've always been told this stuff is really painful, plenty of stories of OEMs sending hundreds of CPUs back for them to work fine once they are returned
@@mrmarecki1 So broadcom nic driver issues have never ever been a problem for you between updates? :D
It's More Common Than You'd Think.
@@Level1Techs Actually, no, and I do work in server space for a large company. If you pull drivers directly from Broadcom issues happen, but likes of Dell and HPE do additional internal validation, so when they release on their websites drivers usually are pretty stable. If you give it a month or two after the release, even more so.
*Me:* Delivered my broken product to cargo officer.
*3 secs later, phone from technical service:*
Your product is working great, so we sent it back to you.
sent*
@@callistoarmy5576 no
@@tf5pZ9H5vcAdBp those* 3 dots
I also make Comparison video please support me
@@tf5pZ9H5vcAdBp did that before I even read your comment.
Just think about all the other common people like us who face such problems but don't have connections like Linus :(
What kind of connections does he have?
@@roskoer5218 Good ones
@@roskoer5218 he means we can’t really speak up like Linus can, cuz he has big audience and we don’t
@@theencryptedpartition4633 cuz Linus is a tech Chad 🤔
Pretty sure "other common people" aren't running business level storage servers?
The little "help" on the drive @5:00 is brilliant 🤣
Linus: Receives a bunch of broken SSDs
Intel to Linus: *That's what you get when you make fun of our chips*
Linus: Intel was actually being generous.
That's true
Did we watch the same vid?
The drives weren't broken and they were also replaced -with more drives that were also not broken.
Linus was using a free engineering sample (aka not for public use/sale) board which was (in this specific use case) not compatible with these drives.
th-cam.com/video/_QXpyNV0bxU/w-d-xo.html..
@@LENZ5369 they should mention something that's not intel fault in the beginning of video
Many viewers they just don't watch until end
"We decided to go with the Dell filled with Liquid Honey Badgers" was a great sentence to be uttered so casually.
@@alinapaasz1841 😂😂
honey badger doesnt give a foooook
yummy
@@ahriik It is a real thing
@@ahriik Liquid is a company and Honeybadger is their PCIe SSD line.
Linus: *Rants about how intel did not do anything*
Linus at the end: oh, its gigabytes issue
It feels more like a Linus issue
Oh look a memory error on the free manufacturer sample server. Ah probably not related. Must be Intel’s fault.
That said this really is an issue with some intel drives.
It is Intel's fault, they gave Gigabyte a faulty driver, that other companies have taken steps to remedy similar cases does not remove Intel's fault
it's*
@@CreeperPookieEither work
The real test would be getting a production model of that server to see if it has the same issue with those drives.
Lol I would never run a company off a Gigabyte server. I would never consider that brand production ready. Non-prod environments, sure, but got to be ready for extra admin management costs involved troubleshooting issues. Real companies who care about up time will be buying Dell, Cisco UCS, HP, Lenovo, and SuperMicro. The fortune 500 company I use to work for would only consider Dell, Cisco UCS, and HP. Working for a smaller company now that hosts everything in IBM Cloud, and all their bare metal server are SuperMicro and Lenovo and I've been impressed so far with them after hearing horror stories from the past about those brands. Linus's use case is almost to the point of buying a real dedicated enterprise all flash Dell or PureStorage array, but half of using DIY is to generate them more revenue with views so reliability and enterprise support & hardware replacement isn't the highest priority.
@@TrevDog513 Dell will probably sell you 4 warranties and 2 antiviruses before they sell you a computer.
If LMG was a regular company:
LMG: Our drives are not working.
Intel/Gigabyte: Is this SSD model on the qualified list for your server type?
LMG: No...
Intel/Gigabyte: OK, thanks, bye.
Of course a regular company wouldn't have a pre-production server to begin with.
yeah, and also if you buy something from ebay without a bill or some sort it's your fault.
These things could also be stolen or something, intel don't has to give warranty to you.
I hate it when some ''influencer'' mean they get special privileges, a normal customer wouldn't had a chance gettinge these repaired/replaced
@@Alainzzz don't know about jurisdiction over there but the warranty should be linked to the product not the customer. It shouldn't matter were you buy a product, if they're under warranty they're under warranty.
@@Alainzzz kinda agree with you, warranty should 100% transfer with a purchase, unless it was already void by previous owner.
But why in the hell does anyone hold a belief that a warranty replacement/RMA would send you anything but the product you purchased.
@@Kilraeus if the product is defective, the producer can choose between issuing an replacement of the same value or more, or they can ask that back and issue a full refund, nothing less. At least it's like this in the EU
"It is possible to commit no mistakes and still lose.
That is not a weakness, that is life."
- Jean-Luc Picard
Check out Git support if you can't commit.
Martin Kase lol
The funny thing is everyone only takes this one way.
If it is possible to commit no mistakes and still lose, then it is also possible to commit many mistakes and still win.
@@PowerStar004 That's not actually a valid logical argument... but the conclusion is true!
@@inv41id Ever played the lottery?
I don’t know what it was about the format of this video but I absolutely love this one in Particular. Love how he’s exposing stuff
Ya but he blamed Intel for most of the video then basically says wait it was the boreds firmware at the end
A tip for when you send something to RMA, before sending it back make a small dot/notch/scratch somwhere hidden so if they send it back you'll know right away.
"Its not broken, you just have to buy the newer model" - Sales Rep
ah, yes, Apple "repair" service
*it's not broken, these shouldn't work in newer cpu especially amd
Yeah I hate that so much. Especially when you vet compatibility on their matrix before you buy things, just to be EXTRA sure, and then the combo isn’t stable you discover after return period and an rma or two.
Also gigabyte support. I got a 1080 g1 gaming and a x470 gaming 7. So they're technically different "generation" parts, different rgb softwares, aaaaaand you can't use one while you have the other. Fucking useful. I love having no lights on my gpu or mobo because of this..... Everyone should just buy Asus.
Thx sir
"You don't need 100Gbps transfer speeds."
That's not what our forefathers said!
100 GB/s*
no they said“640K will be enough for anyone” ( ibm 1981) lol
@@dr_crimbo who could ever fill up 1gb of storage?!
@@MisterRorschach90 never, it's impossible
I really enjoyed this and watching you get into that troubleshooting zone. Makes me miss working with physical systems, even when it doesn't go well
Here in my country, when you send something to the company for replacement, you put a word with a sticker, or a dot of paint, inside the product or in a place difficult to notice at first sight. When it comes back, if it still has the paint inside, you know you're screwed... Works extremely fine with car parts, but also with computers and other things.
This video had more twists and turns than my cable management
Edit: thanks for all the love peeps! 1.2k likes!
Good one
It's more bent than me
@@JTS-Games agreed
Damn, I don't even have any
Post popularity edits are still cringe btw
11:35 Linus really pulling a Vsauce
That needs the typical 'dong' that comes after that sentence.
@@haydenrogers3486 Yea needs an edit with the Vsauce music lol
lmao was thinking the same thing... he also did it at 5:49 which made it feel even more vsaucey
Oh man I forgot that channel exists
HAHAHAHA well spotted sir!
I love this sort of deeply nerdy server troubleshooting video. I geek out hard for this
Even if these videos don't turn out as intended, I do find them nice since they dive into more of the niche topics on storage at this sort of scale.
@Prie Boemee
Stop trying to steal from people.
Go away.
8:36 - "Hey, you guys want a free tech tip?"
Yes Linus. That's literally why I'm here.
Could I get a few of these broken SSDs instead maybe?
"It would cost the same to send a product that works"
- Me sitting on my 5th Razer Basilisk showing signs of double clicking asking for a replacement with optical switches.
Razer mice always have those problems.
Take. It apart and dump rubbing 99% alcohol down into the switch click it a bunch then wait 2 hours. Magic
@@djscottdog1 My last one (a OG Abyssus) is 10 years old and isn't double clicking.
And, when I search this problem, I see more threads from people with 502s.
As far as I can tell, the problem is less Razer and more these particular Omron switches.
Which, I suppose does make it Razer's fault for using them, but it seems everyone's using the same shitty switches.
My Naga Trinity is doing it too.
@DJ Scottdog Logitech mice have the same exact issue and both razer and Logitech solved the issue on there newest mice.
4:50 that's a nice intro to polls vs interrupts for everyone actually
5:00 that simple "help" just made my day
I was just having fun listening to the server fans gradually spool up while Linus was talking.
A robot speaking needs some power
I dont hear fan noise anymore.
My work has so much of that I learned to ignore it.
@@MrMarci878 xxxxxzzx
The person above this comment is joshua b... and youre joshua c...
@@MrWestonO Hah, I didn't notice that! that's cool.
The little "Help." from the SSD was gold. Like, it's scared of the big hairy human yelling at it. xD
The new dark theme intro Is GREAT, i like it.
Three emotions from 0:47 to 0:49
Excitement, shocked, disappointment.
Lol Linus you are a true actor
lol true
😲
That's just how good their sponsor, smart deploy is.
Smart deploy enables IT admi.......
th-cam.com/video/_QXpyNV0bxU/w-d-xo.html..
@@klaudiajenka3987 😍😍
Have a couple of these servers, I needed to enable IOMMU to prevent drops.
My mom bought a couch. It came broke so we sent it back. They sent back the same broken couch back about 7 times before we got a proper fixed one.
You got to admire how resolute the guy at the RMA department must be. Sending it back once and then giving up? Nah. Giving up after sending it back only 5 times without doing jack shit about the problem? That's just quitter talk. This guy sent the SAME broken couch, which he could, I assume, diagnose with his own two eyes SEVEN times!!
@@fredwupkensoppel8949 löl
@@fredwupkensoppel8949 RMA Dude: "Can't let'em win this battle. Send it back to them..." RMA Dude's Underling: "A new one, sir?" RMA Dude: "Are ye fookin' kiddin? Send the one they sent to us back to them!!!" RMA Dude's Underling: "You're a wise one, sir! Right away, sir!"
5:02 Wendell is indeed some kind of genius. Brilliant guy
My bet: linus wants to subsidize the cost of new drives making a video about the broken ones
Linus wants to afford some SSD brand's marketing department the opportunity to sponsor an update video* ;)
that's kind of the whole point of LTT Media Group
That sounds right, this video was just him whining
I was expecting for Linus to say “see, they still have the drop marks from me”
This is pretty high on the list of problems I can't relate too!
But great content, fun to watch guys!
such smooth transitions to sponsors everytime gets me!
Original clickbait title for when they inevitably change it: They sent my broken product back.
No one cares
2 things I don’t understand in life
1: Mathematics
2: *Why people don’t give me a chance to make it on TH-cam even though I work so hard making good content*
Help
@@eugenesmemes stfu, grow on your own
thanks for the likes, ily all 💕
@@nemiesis No one cares about you either.
@@eugenesmemes 1. You complain on random yt videos 2. You only record short clips of a medicore game
You got another video for me yet?
What about now?
You got one now?
Maybe now?
What about... *now?*
no not yet piss off
@@hughlevantjames905 calm down it's a reference to something in the video
It's a pretty good joke ngl
And it just went over your head
@@grod5998 no... it didn't go over my head. I was just continuing the joke, thanks for ruining it dumb**s.
ye it's actually good, much better than the generic overused templates for _everything_ in yt
@@hughlevantjames905 that was a joke? Damn bro u got the whole squad laughing 😐😐😐
This is a very good troubleshooting lesson that can be used in so many different applications.
Segue Grade: A+ Did not see that ad coming, great work!
I feel like the title is kind of unfair after having watched the video.
"They sent back the same incompatible product"
would be more fair, as it appears to only be an incompability issue and not an actually broken product.
Yeah, and starting the video off on the wrong foot by blaming Intel knowing full well that many viewers won't watch until the end is quite disingenuous. I feel as though there should have been a quick "future Linus" interruption to explain that it didn't quite turn out to be Intel's fault.
I think Linus was expecting at least a firmware update to resolve the issue
The fact that it was likely Linus' use of a pre-production rack from Gigabyte that caused the issues only makes this title more laughable
I feel like buying second hand drives and then expecting to get new drives from intel because of an incompatibility issue, to the point you are leveraging your contacts is pretty arrogant.
@@mysterythin I don't understand why they were RMAd, just work with Intel on a firmware patch... this is a brain-dead RMA
My guess is that the drive firmware is actually programmed to send the interrupt a little early to compensate for interrupt latency on old systems. The problem would then be that newer systems respond faster. Have dozens of interrupts arriving early at once, and the CPU load from polling drives that aren't actually ready yet starts causing stuff to time out.
Doesn't quite explain why a single drive seemed to work fine with no issues, only when multiple drives are being loaded.
@@crenn6977 Easy! With a single drive, the polling only consumes a little CPU time, so it doesn't time out. With many drives, the interrupts stack up and start causing more and more delays until things time out.
4:53 closed captioning (pull the drives)
I believe that was POLL the drives. Thats a computing function.
Pulling drives is a human function, physically removing them.
I love when he roasts any company’s customer service
11:30 hey vsauce Michal here
the "help" on the drive as linus asks it if it got something for him, man that got me good
While I understand the reasoning behind clickbaity titles, I actually prefer to know the contents of the video before watching. In this specific case I feel like the title, considering the content, was unfair to Intel - they sent you a replacement for a functional product due to an issue that is probably not their responsibility.
true
From what I understood of the issue (the drive was triggering an interrupt before it was ready) it seems the issue really was on Intel's side and they should have fixed it (probably easy enough to do with a firmware update). The fix the other MB manufactures made basically overrode the standard behavior for this one particular drive but wasn't really their responsibility (they did it most likely because it was enough of a problem for them to do so). In their defense, as Linus pointed out, the issue only occurred with certain hardware and configurations that weren't too common but they still should have fixed it. I imagine the reason they sent the interrupt when they did was to compensate for the response times of other processors thereby creating a performance boost by risking this very issue.
As far as clickbaity titles go, unfortunately that's just how it works. Linus has stated in the past that he would much prefer to not use the clickbaity titles but statistics show time and again that clickbait sells so as a business owner he has to do, what he has to do.
It's linus, you should be used to clickbait and low effort. Where do you think you are? Gamers Nexus?
Intel seems to have broken spec from what i understand. That means it is their responsibility.
@@lisafairymeadow2034 If you think his stuff is low effort why are you even here? Go somewhere else instead of hating on content.
I love that automatic CC's have captioned this as parody mode. It's actually apt here.
Linus: "Hey guys want a tech tip?"
Me who thought that was the name of the channel:
I love the "Here's a free tech tip for you". It feels like a half ironic take on the channel name, but is also interesting toO!
I love simple videos like this where Linus just rants about companies in his office XD
The title of the video: "They sent my broken product back."
Isn't exactly consistent with the content in the video
Yup and it proves Linus can be a big ole douche sometimes. All he did was bitch about intel and never once said he was in the wrong. Even at the end he said “thanks intel I think”
@@kman1898 but he... Wasn't in the wrong?
Don't get me wrong, this was just an RMA miscommunication, intel asked no questions and linus told them nothing, but linus isn't wrong for expecting to get a working product back, and intel wasn't wrong to send them a working product back.
Linus made clear it was a very niche edge case that caused the problem.
To say linus is somehow in the wrong here is pretty silly, when the only thing wrong they did was not communicate better with the RMA people about what they wanted, and considering the size of the organization, and the fact that flunky number 2 said "they asked no questions" it seems like everyone expected different things without making sure that's what they were gonna get.
You're right, the title is a bit silly when the drives aren't technically broken, or even the same drives, but your rudeness is a bit uncalled for.
@Draw a Pretzel no, Linus was definitely in the wrong here. Linus is also way more rude because of what this video insinuates which has weight and can affect sales by hurting intel's brand image. And this proves LMG has gone to the shit with all the clickbait style titles and thumbnail captions. I'm sorry but if someone is being an entitled brat about not getting what they want, don't expect people to respect said person. Linus was in the wrong here for expecting an engineering sample from gigabyte to work in a production environment as if LMG can't afford to buy one that's guaranteed to actually work by contract. As a huge influencer his words have weight and this can/will negatively affect Intel because of how irresponsible and unprofessional they are.
@@drawapretzel6003 To make it clearer than Darrin did... In this case LMG has no case because they purchased the drives second-hand and that basically always means the warranty is null and void. Warranties for most everything only apply to the original purchaser and nobody else.
That's why Linus had to go through Intel's PR department to get ANY support because it was basically: help me or I'll raise a stink and you guys won't like the bad publicity. But then because they didn't replace the drives with the latest hardware (new generation that doesn't have this issue), he's raising a stink anyway.
Not as strongly, but it's still there.
@@drawapretzel6003 Linus is wrong because generally you don't get your money back or a completely different product in return just because you happened to order the wrong product on eBay, not realizing it won't work in your setup.
As much as I despite Intel and refuse to buy from them, most companies don't let you return products for whatever you like just cause you fucked up, and Linus doesn't deserve special treatment just because he's a TH-camr. Warranties don't tend to cover any of this, and RMA's certainly don't (because the product isn't necessarily fault here).
Yes Linus... I said it! Our secret weapon "Get Yvonne to yell at them" lol 😆🤪😄
Imagine being an ordinary man, not having an insanely popular TH-cam channel where you can report to the public how a retailer and/or a manufacturer has screwed you over.
Send a formal notice
LMG is 100% in the wrong here. I doubt Intel SSD model was qualified for Gigabyte AMD-based engineering sample server, so I'm pretty sure their warranty claim should've been instantly rejected. If you try stuff like this, it's at your own risk. Intel replacing drives at all was already a PR good will gesture.
@@mrmarecki1 Yeah, servers are picky about anything you put in them. I have an Dell T130, and it only works with Unregistered Ram from Samsung. Not regular and not registered Ram, and only from samsung.
th-cam.com/video/_QXpyNV0bxU/w-d-xo.html..
@@mrmarecki1 I work at intel in customer board development and I can almost guarantee these weren't validated on a platform like this, like you say, this is entirely a gigabite /AMD compatibility issue
Been watching for about a year now. I still have no idea what is going on 90% of the time... but I hope I am learning something for when I plan to build a new pc in a few months lol... keep up the good work team
Lol those two sponsor segways were possibly the best segways I have seen you do on any video so far. 😂😂😂
"You ever spent a lot of money on something?" -Sees a golden Xbox controller- Not like that but yes.
They little "help" on the drive literally killed me lol
Dude!! I'm going through that same exact thing right now with my Asus ROG Zephyrus M15 GU502GV!! I had accidentally let it die a few weeks ago and just put it on the shelf and didn't touch it for 2 days. When I went to use it (to clean it out to sell) it wouldn't turn on and no power light when the charger was plugged in. I bought it May 5th of 2020 and my warranty was about up. So I sent it back and they said I needed a new motherboard and that they replaced it and sent it back in perfect working condition. When I got it back, absolutely nothing happened. Same exact problem. So I had to send it back to them again, and now I'm waiting for them to repair it and send it back for the second time.
I actually have a habit of disabling some intel options in my bios or changing intel divers back to generic ones in windows. All due to stability or bugs with intel software on some devices, like intel sound driver issues on lenovo laptops (brand new ones).
Most Intel 'features' and 'options' are critical vulberabilities used to root your system anyway. Malware using Intel ME is literally Ring -3 and owns YOUR system.
Hi Linus!! Huge fan. Long time follower. Have a great rest of your day :)
😩
@@callistoarmy5576 no ❤️
I havent been this early since windows xp came out
2 things I don’t understand in life
1: Mathematics
2: *Why people don’t give me a chance to make it on TH-cam even though I work so hard making good content*
He1p me!!!
Same here
@@eugenesmemes pls stop
I haven't been this late since voodoo 3 came out and I found out it didn't work with windows xp
@@eugenesmemes Do you thing everybody makes big on TH-cam after only posting videos for 2 months
Would be interesting to put an oscilloscope on the power rails going into that backplane and pins on the controller and see if you are browning anything out under load.
Very good point! Way back in 486 days our high end server board ran fine until hit with heavy continuous interrupts. Today dozens of USB drive adapters crap out when writes queue up. Things like internal ground bounce do not show up on a scope, but you first have to check the backplane as you said. A backplane that works with a full set of x-nm drive controllers may not work when loaded up with x-1 nm drives. Kudos if you have a GHz scope that can check that. If both are marginal, who gets the blame? Seen on a compatibility list : "only one of these per system is supported".
I hope y’all can fix this issue with zen 3. I tried to find a few of these online when y’all made your first video. But there were none. I just need 4 ish. I would use these in my work station as scratch drives. Im trying to edit 8k footage. I’ve been looking at purchasing the sabrent rocket pcie 4.0 4TB.
Anyway, i need a gpu first, since i color correct with davinci resolve. So ive got a couple years before ill be able to upgrade.
I look forward to each of your server videos. Love this one!
I feel for you. This kind of small issue is the kind of thing I run into with my home lab all the time. I bought a bunch of SLC SSDs from the early days for a raid 10 and get occasional random write errors. Heat and EMI seem to be a factor. The thing is, this is what happens when you buy a part outside of normal channels for a use case that hasnt been tested.
This is why I don't bother trying to deal with companies anymore. Product arrives broken or defective, I immediately return it and let the retailer eat the cost. This happens enough and it will sour the retailer/vendor relationship.
Costco is great for doing this. They basically will take anything back.
Ah yes, the rare armchair tech diagnosis in the wild. Please tread carefully, they're very easily triggered by mundane events.
welcome to my world linus! in my life i have already sent some defective products back to various manufacturers as a guarantee case, and several times I have received back the same defective product that I have complained about as defective. sometimes you have the feeling that many manufacturers or dealers are just idiots in the complaint departments. greets from germany
Love the tech videos like this one =)
Wait he purchased drives from eBay? I don't know of any manufacturer that would honor a warranty from a drive purchased from eBay...
You think Intel is bad? I've seen the same unrepaired part come back on nuclear reactor I&C equipment. Same serial numbers, same defects, same "bad" sharpied on the outer box.
I feel cold all of a sudden
@@Yonix06 don’t worry, you’ll feel _very warm_ quite soon…
@@randombloke82 Your reply just made my day.
A couple of years ago I too experienced a bad combo of Intel hardware. It was Intel's 520 SSDs (with a SandForce chipset) and Intel's own laptop motherboards HP were putting into their early 11-inch x360 laptops. I was upgrading my sister's laptop from a spindle. What I recall was that the SSDs would never boot up once they had gone into low-power mode. This meant they operated fine ... *once*. After installing Windows, rebooting the laptop would give you a system with no drive found. You had to unplug the internal battery in order to boot again. It took a while to figure out the incompatibility. I was able to confirm that the problem *only* happened with that specific motherboard chipset and that specific SSD chipset. Except for the wasted hours spent troubleshooting, it did at least end okay. I happened to have purchased a different non-Intel SSD at the time for another spindle-to-SSD upgrade ; so I just swapped them around. :)
Same what was happening to my Firecuda sshd last week. Would drop out even with a slightest read/write load. Just because its ssd part became so full that it was performing worse than a 2012 hdd. Its has a fundamental flaw in how it handles files on the ssd part.
3:14 Just FYI guys, before we had Anthony, we had Wendell to answer the tough questions. And apparently, there are still questions only the man himself can answer.
(No hate on either one, I love both of our geniuses)
and when it comes to storage stuff, i actually see Jake more often than Anthony though
wendell would make anthony feel dumb and i doubt he'd have any trouble admitting it, wendell make most of us feel dumb
@@JohnSmith-us4pj right!? Remember back in the day, he had Logan to dumb it down for us, so we wouldn't feel like complete neanderthals. But now with lvl1techs pfff...
12:15 Immediately the higher pitched noise makes me hopefull that this server will be better. More vrrrm more good, no?
If you guys dont want that RAID5/6 write performance hit in Windows storage spaces, you have to align the interleave value of storage spaces with the block size of the filesystem. Example: If you use 5 disks for RAID 5, set the interleave to 16KB and the NTFS/REFS block size to 4x16KB=64KB. The fifth drive is just parity ... Another example would be RAID6 - 8 drives with 16KB interleave and NTFS with 128kb bloack size. In both cases the writes work ... most of the time, its windows after all, it cant be bothered with consistency and rules.
PS cmds for that stuff:
New-VirtualDisk -StoragePoolFriendlyName "Home_5x1TB_Parity" -FriendlyName "Home_5x1TB_Parity" -ResiliencySettingName Parity -PhysicalDiskRedundancy 1 -ProvisioningType Fixed -UseMaximumSize -NumberOfColumns 5 -Interleave 16kb -Verbose
New-VirtualDisk -StoragePoolFriendlyName "Home_Storage_Pool" -FriendlyName "10x1TB_DualParity" -UseMaximumSize -ProvisioningType Fixed -ResiliencySettingName "Parity" -PhysicalDiskRedundancy 2 -Interleave 16kb -NumberOfColumns 10 -WriteCacheSize 0 -AllocationUnitSize 256MB -NumberOfDataCopies 1
I have exactly the same feeling and situation.
"Disk x in Backplane 1 of Integrated RAID Controller 1 was reset."
"Fault detected on drive x in disk drive bay 1."
It's not a disk fault, at least before it inserts into Dell server.
I don't know if I'm alone, but my second hand DELL 730XD+ESXi 7 with PERC 730P Mini for home lab is keeping resetting ALL the disk (including WD/Seagate/Toshiba HDD, Toshiba/Liteon/Sandisk/Others SSD) randomly, during any workload, for no reason, causing VM to freeze or crash.
Sometimes the SSD will be "locked up" or "READ ONLY", and doesn't work without secure erase (pulling out the disk, use other laptops with Pmagic), the Sandisk Ultra 4T from Amazon even "Killed" immediately before "online" (haven't even created a partition table yet, even secure erase can't recover that, current in RMA.
I've changed a controller card from the seller, try fully upgraded firmware and drivers, even downgrade to try the latest 3 versions of them, not work.
Some of them are in raid-0, some of them are not, and the HDD (in China) has been crazy because of the Chia coin, even can't buy a 10T up in normal price for migrating to non-raid. (by the way, when the reset occur, all disk include non-raid are reset)
After some search, I found Dell Controller (maybe) does not happy with third-party parts, especially Samsung SSD (ok, not my case), LiteOn (some FW bugs?), in case SSD cause problems, I've removed all SATA SSD now. The new "native" driver for ESXi (lsi-mr3) has a bad reputation, MANY OFFICIAL docs show "advance" but "required" tunning configure (kb.vmware.com/s/article/2144936) need to be set to avoid problems, one official KB even tell you to disable the new driver and use the old one as a "Best practices" (kb.vmware.com/s/article/2109665). But there is no other choice in ESXi 7, they completely remove the vmkernel driver. and All seem useful tips from the community and docs are tried, increase io timeout, increase queue, factory reset, disable r/w cache, balabala, even downgrade the ESXi itself from 7.0 U2 to 7.0.
One thing that didn't change is the backplane, I don't want to bother the seller anymore, and the front and back disk are both resetting, using different cables to MB, so (I think) it's very little change a backplane problem. (tell me if it has a chance)
I know, Dell/VMware could have some explain, like your disk is not Dell Part, or not on HCL, but at least, it's using new enterprise-class HDD (HC320, HC550).
Yes, VMware was acquired by Dell. But not works well with Dell's controller, do you believe that?
Not to mention other manufacturers (datacentersupport.lenovo.com/us/en/products/servers/thinksystem/sr950/solutions/ht505543-vmware-local-datastore-disconnect-and-virtual-resets-on-the-logs-lenovo-thinksystem-and-lenovo-server), "There is no information from VMware about the reason for the VRRs, even though the storage devices were working optimally." (I don't know if they are linked problem, but seems.)
It's very annoying, now my troubleshooting route is to wait for a normal price of HDD for migrating to non-raid, change to an H330 HBA card, downgrade to ESXi 6.5 to try the megaraid driver instead of "native" lsi_mr3. If all the above does not work, I may change to a quanta server, with no "Dell Part Number" HDD needed, maybe keeps the problem away.
Nothing will describe my disappointment about Dell.
This video was a rollercoaster of emotions
10:28
Big Pink Bottle but the bottle is grey
Looks like the sad filter was applied on the LTTStore too
Saw that, but linus bottle was a large pink one thought
Man managing server farms and all the costs and headaches that come along with it is one thing I do not envy
Linus: They are second hand but are still within warranty.
Intel: It's only the original buyer that has a warranty.
Another thing to consider would be whether the drives were retail or OEM. With OEM drives warranty and tech support are handle by the system manufacturer not the part manufacturer. If these were OEM drives someone pulled from a server, Intel probably isn't obligated to warranty them, especially if they work just fine in the original equipment they were installed in. The fact they might not work in some pre-production beta piece of hardware from a company like Gigabyte, would be moot.
I've had to RMA my Gigabyte 3080 now I'm seeing this video. Linus you're terrifying me.
I have rma a motherboard before don't worry it should be fine the reason this didn't work for him it's because the issue is so specific to reproduce
I know this isn’t exactly the same but I was given a burger with cheese on it even though I specified I didn’t want cheese because I had a dairy allergy and surprise surprise I gotten cheese on my burger... so I took it back and they offered a full refund or some food instead (it was took back and they were told it was took back because it had cheese on it and I am allergic to dairy) and the food offered was a McFlurry...
I was reading being sad you didn't mention the specific restaurant, but the last word gave me my answer hah
Telling you about our Sponsor... "Intel..."
The best videos are the ones where linus starts to foam at the mouth just a little 5:24
This video was fascinating! Such a strange edge case.
So this video was "I'm gonna yell at Intel the whole time when... it's not their fault"
*It's not just their fault ...
A firmware update from either Intel or Gigabyte could have fixed this, but neither found it significant enough to address
@@Dream0Asylum Likely it'd have required a hardware change to fix on Intel's side. But yeah, if they could address it via a firmware update, they should have. But that's why I doubt they could have since they're not dumb and know how much hassle this can be for large customers.
For Gigabyte, they clearly could have fixed it via a firmware update and should have since the patch was already written and available to them by AMD. They just didn't do it. So the blame is more on them than Intel in this case. Which was my point.
Also, the drives are not covered by a warranty any longer because warranties typically do not transfer between users. That's why Linus had to go through Intel's PR to get any support since he had purchased them second-hand. And then he proceeded to make a video complaining about the support they tried to give...
In dealing with issues like this at work, the RMA people usually have a set procedure they execute which is simply to send replacements of the same model unless other arrangements have been specifically made. The reason they do this is to try to not introduce another incompatibility in the system, such as if you were only replacing one or two drives (ideally you'd want all of them to match). Unfortunately I've been in a position where you get vendors like Intel and Gigabyte pointing fingers at each other as to who is at fault, hoping your warranty expires before any traction on the issue occurs.
There's no way in hell I'd spend $10 let alone $10k for used shit on eBay and not expect various levels of bullshit to deal with. It's why I avoid "bargains".
I got recommended that video 5 hours ago, and now there's a continuation. What a service!
Algorithm be smart. These videos are all scheduled. I wonder if it was recommended because it recognised it was a sequel.
@@xenofenus Must have been it
When I was in college I interned with Intel and my roommates were with NSG (Nonvolatile memory Solutions Group). They flat out told me to never buy an Intel SSD.
Why tho
@@leonardoanchundia2721 They're not as good as the competition. You'll get better components, better QA, ect by going somewhere else.
This is why I will never purchase another product from EVGA. I bought an OG Titan that failed. Sent it in for RMA, they suckered me out of another $45 for the advanced RMA program before sending a replacement - then the replacement had bad VRAM
I sent a faulty CD/Radio stereo system back some years ago and so I wouldn't get the same one back as a replacement I copied all the numbers like SN to a word document printed it and put it in the box to return it. They sent the same one back, it even had the paper I put in with the SN and other info on it in the box. I contacted them by phone and they said I had to send it back at my expense AGAIN so I told them I was initiating court action that day and they quickly said they would send a new stereo and pick the old one up at the same time, it took them 21 days to deliver and pick up the old unit.
I love how he is always so smooth about going into the sponsor even with the sharp turns