Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
eviltastic
Feb 8, 2004

Fan of Britches

well why not posted:

Jay’s old sidekick went by “coconutmonkey”. Do with this what you will.

If there’s fault to find there, it’s probably with PC Gamer magazine. That’s the name of their mascot since the 90s, and they reference it all the time in their community engagement, so I’d think that’s the reference.

Adbot
ADBOT LOVES YOU

CoolCab
Apr 17, 2005

glem
also i'm pretty sure that's the dude jay has brought back recently and if so he's Black and presumably chose that name on his own, so idk how you evaluate that there?

jay sucks but he puts up consistent content. :shrug: i kind of like watching his "here's a problem we're trying to fix" style videos for interesting troubleshooting but most of his stuff is very dire.

Koskun
Apr 20, 2004
I worship the ground NinjaPablo walks on
Over the past few months Jay has said he isn't quite sure what videos to keep making, hence why there has been so many talking head ones lately.

He's as frustrated over the video card shortage and that every launch is seemingly a paper one. Very annoyed at nVidia and AMD for releasing the ti/XT line when they can't keep up with the normal versions as well.

Out side of the oddball cases he seems to get sent semi-regularly (which is his latest personal build), there isn't much to review for a site like his it seems. LTT has a broad array of people with different hobbies and such, so they hit a lot of things. For Jay it's him. There are only so many ways to do a "custom water loop", which he's acknowledged too. He did some other stuff last year (the drone stuff was amusing if nothing else), but yeah, how many ways can you watch a computer being built?

Endymion FRS MK1
Oct 29, 2011

I don't know what this thing is, and I don't care. I'm just tired of seeing your stupid newbie av from 2011.

Koskun posted:

Over the past few months Jay has said he isn't quite sure what videos to keep making, hence why there has been so many talking head ones lately.

He's as frustrated over the video card shortage and that every launch is seemingly a paper one. Very annoyed at nVidia and AMD for releasing the ti/XT line when they can't keep up with the normal versions as well.

Out side of the oddball cases he seems to get sent semi-regularly (which is his latest personal build), there isn't much to review for a site like his it seems. LTT has a broad array of people with different hobbies and such, so they hit a lot of things. For Jay it's him. There are only so many ways to do a "custom water loop", which he's acknowledged too. He did some other stuff last year (the drone stuff was amusing if nothing else), but yeah, how many ways can you watch a computer being built?

I’ve noticed myself watching FAR less tech content precisely because of this. Who cares about reviews when I can’t even buy it. It’s even trickled down to other components, I couldn’t tell you how good or bad the 12th gen Core stuff is because when I saw the reviews I just thought “who cares”. I’d love to get back in but I’ve even flirted with the idea of swapping my PC for a MacBook because I noticed all I ever play are indie/retro/generally less demanding stuff

Pilfered Pallbearers
Aug 2, 2007

Endymion FRS MK1 posted:

I’ve noticed myself watching FAR less tech content precisely because of this. Who cares about reviews when I can’t even buy it. It’s even trickled down to other components, I couldn’t tell you how good or bad the 12th gen Core stuff is because when I saw the reviews I just thought “who cares”. I’d love to get back in but I’ve even flirted with the idea of swapping my PC for a MacBook because I noticed all I ever play are indie/retro/generally less demanding stuff

TBF the only things in shortage right now are GPUs and DDR5. Everything else is easily available.

njsykora
Jan 23, 2012

Robots confuse squirrels.


GN’s talked about that video struggle as well, no-one’s looking for GPU reviews when they can’t buy them so you have to start branching out into other stuff. Like prebuilts.

Arivia
Mar 17, 2011

njsykora posted:

GN’s talked about that video struggle as well, no-one’s looking for GPU reviews when they can’t buy them so you have to start branching out into other stuff. Like prebuilts.

thankfully, steve and team are really good at making anything remotely in his wheelhouse fun and interesting without cheap or stupid gimmicks. moving to the new place has also given them a bunch of video content, from moving vlogs to stuff the new place enables, like the fan tester.

Endymion FRS MK1
Oct 29, 2011

I don't know what this thing is, and I don't care. I'm just tired of seeing your stupid newbie av from 2011.

Pilfered Pallbearers posted:

TBF the only things in shortage right now are GPUs and DDR5. Everything else is easily available.

I know. It’s just the GPU shortage causing me to be apathetic overall

Koskun
Apr 20, 2004
I worship the ground NinjaPablo walks on

Pilfered Pallbearers posted:

TBF the only things in shortage right now are GPUs and DDR5. Everything else is easily available.

Which isn't really fair at all. Want to build a computer, well hope you like onboard graphics. Want to build a new 12th gen Intel, hope you want no upgrade path because you have to resort to DDR4.

That doesn't even touch the shipping issues of getting something back in stock. What might be "easily available" today can be gone with no ETA next week pretty easily right now.

Sure, if you want to pay scalper prices, anything is available, to be fair that is.

Pre-built is honestly about the only way to get a GPU, and probably DDR5, now a days at anything close to MSRP. For most people that works, however how many times can you review what are essentially cookie cutter computers as well?

Koskun fucked around with this message at 01:12 on Feb 3, 2022

Pilfered Pallbearers
Aug 2, 2007

Koskun posted:

Which isn't really fair at all. Want to build a computer, well hope you like onboard graphics. Want to build a new 12th gen Intel, hope you want no upgrade path because you have to resort to DDR4.

That doesn't even touch the shipping issues of getting something back in stock. What might be "easily available" today can be gone with no ETA next week pretty easily right now.

Sure, if you want to pay scalper prices, anything is available, to be fair that is.

Pre-built is honestly about the only way to get a GPU, and probably DDR5, now a days at anything close to MSRP. For most people that works, however how many times can you review what are essentially cookie cutter computers as well?

Nearly everything outside of GPUs and DDR5 are available, easily, for MSRP. Even intel 12th gen chips and associated boards.

We’ve also not been recommending DDR5 in the PC building thread. The only DDR5 that’s not basically the same performance as DDR4 has an MSRP of $690 for 16GB.

The prebuilt stuff is true, specifically for obtaining a GPU only though.

EngineerJoe
Aug 8, 2004
-=whore=-



well why not posted:

Jay’s old sidekick went by “coconutmonkey”. Do with this what you will.

Named after the old PC Gamer character?

https://www.youtube.com/watch?v=J1Xf53ljCMg

lurker2006
Jul 30, 2019

njsykora posted:

GN’s talked about that video struggle as well, no-one’s looking for GPU reviews when they can’t buy them so you have to start branching out into other stuff. Like prebuilts.
You know I have to wonder how bad the gpu shortage actually is for tech content creators. You'd think normally the average person would be plugged into the general pc consumer sphere for a bit while they research their builds, but after they exhaust their bank account for the build and put it together successfully, with nothing else to buy or tinker with, they'd drop out of the viewership. Now you have people locked into it for months waiting for a gpu so they can start a new build, "hell why not watch another case review, maybe I'll get that one."

Klyith
Aug 3, 2007

GBS Pledge Week

lurker2006 posted:

You know I have to wonder how bad the gpu shortage actually is for tech content creators. You'd think normally the average person would be plugged into the general pc consumer sphere for a bit while they research their builds, but after they exhaust their bank account for the build and put it together successfully, with nothing else to buy or tinker with, they'd drop out of the viewership. Now you have people locked into it for months waiting for a gpu so they can start a new build, "hell why not watch another case review, maybe I'll get that one."

It's a nice theory, but that isn't happening. All of them are explicitly saying that viewership & hits are down on content that was their bread & butter. Go back a few pages to the twitter convo between GN Steve and Ryan Short of Anandtech. They disagree what the best solution is, but they both say that it's hurting them.


I do think there are factors besides the GPU shortage:
• New consoles came out a year ago. That draws off the audience who aren't dedicated PC enthusiasts for a while. If you have a shiny new PS5 or XSX you're looking at console game coverage, not PC parts.

• 2021 was a down year in (AAA) game releases, lotta delays and weak releases. The question "should I build a new PC just for Battlefield 2042?" had a pretty easy answer. The breakout hit this year was Valheim, which is pretty middle-of-the-road in terms of how much grunt you need to run it.

• Hangover from 2020. A lot of people built PCs in 2020. Sales were way up despite the everything shortage. So most of the people who could afford overpriced gear are done and not paying close attention anymore.

gradenko_2000
Oct 5, 2010

HELL SERPENT
Lipstick Apathy

Klyith posted:

• Hangover from 2020. A lot of people built PCs in 2020. Sales were way up despite the everything shortage. So most of the people who could afford overpriced gear are done and not paying close attention anymore.

I did a review just now of what I did in 2021 and it was moving to an i5-10400 in April, getting moving from 16 to 32 GB of RAM in June, and then nothing else for the rest of the year. I actually bought an LGA 1200 board but I never bothered picking up a CPU for it. I guess it's pretty true - Zen 3 always felt too expensive, Intel 11th gen was only an incremental upgrade, and I'm just waiting on DDR5 to get more accessible even if 12th gen is a big enough leap to be kinda interesting.

Former Human
Oct 15, 2001

A Linus "we lost all our data" video must naturally be followed up with a "look at our NEW server (which is basically the same as the last one that failed)" video.

https://www.youtube.com/watch?v=pqZFcrQTZ_E&t=257s

FYI they spent $36,000 on Seagate drives that they bought from a third party seller off Newegg. Any bets on how long this one lasts?

I would blow Dane Cook
Dec 26, 2008

BlankSystemDaemon
Mar 13, 2009



Former Human posted:

A Linus "we lost all our data" video must naturally be followed up with a "look at our NEW server (which is basically the same as the last one that failed)" video.

https://www.youtube.com/watch?v=pqZFcrQTZ_E&t=257s

FYI they spent $36,000 on Seagate drives that they bought from a third party seller off Newegg. Any bets on how long this one lasts?
Haven't watched the video, and i'm not going to - but that's probably some of the poorest choices they could've made:

(Source)

mobby_6kl
Aug 9, 2009

by Fluffdaddy

gradenko_2000 posted:

I did a review just now of what I did in 2021 and it was moving to an i5-10400 in April, getting moving from 16 to 32 GB of RAM in June, and then nothing else for the rest of the year. I actually bought an LGA 1200 board but I never bothered picking up a CPU for it. I guess it's pretty true - Zen 3 always felt too expensive, Intel 11th gen was only an incremental upgrade, and I'm just waiting on DDR5 to get more accessible even if 12th gen is a big enough leap to be kinda interesting.

I did nothing in 2021. In fact my PC is literally untouched since I got the 1070.

I'm doing a bit more video editing now though so I was also thinking of getting alder lake. DDR5 would be great though.


BlankSystemDaemon posted:

Haven't watched the video, and i'm not going to - but that's probably some of the poorest choices they could've made:

(Source)
Linus lol.

power crystals
Jun 6, 2007

Who wants a belly rub??

Former Human posted:

FYI they spent $36,000 on Seagate drives that they bought from a third party seller off Newegg. Any bets on how long this one lasts?

I think Seagate gave them the drives and they were just using that third party seller to determine the market value since nobody else had any yet.

Not that it makes any of the rest of it less stupid.

well why not
Feb 10, 2009




yeah he says in the video that seagate sent them

Devian666
Aug 20, 2008

Take some advice Chris.

Fun Shoe

well why not posted:

yeah he says in the video that seagate sent them

I guess they were running low on 14TB drives to send him.

Combat Pretzel
Jun 23, 2004

No, seriously... what kurds?!
The Seagate STxx000NM the same as the STxx000VN drives? Because I got the 10TB Ironwolves for my NAS. :|

SlowBloke
Aug 14, 2017

Combat Pretzel posted:

The Seagate STxx000NM the same as the STxx000VN drives? Because I got the 10TB Ironwolves for my NAS. :|

He got a crate of exos, the equivalent of wd golds. Ironwolf are a lower tier in their catalog.

Klyith
Aug 3, 2007

GBS Pledge Week

Combat Pretzel posted:

The Seagate STxx000NM the same as the STxx000VN drives? Because I got the 10TB Ironwolves for my NAS. :|

%1-%2 AFR is the overall average for their drives. I'm not sure why BSD found that model to be lol-worthy, because it seems perfectly normal.

They're breaking out 24 different models, so by random variation some will be higher and some lower. The more groups you look at, the higher the expected variation. Likewise, there's a drive on that set of charts with almost 5% AFR -- but only for 2021, versus in 2020 it had 0%.

If you play around with subgroup statistics enough, you can show that horse dewormer is a cure for covid.



You want to see a defective drives, look back at the old blogs from 2015-2017 and check the seagate failure rates.

BlankSystemDaemon
Mar 13, 2009



Klyith posted:

%1-%2 AFR is the overall average for their drives. I'm not sure why BSD found that model to be lol-worthy, because it seems perfectly normal.

They're breaking out 24 different models, so by random variation some will be higher and some lower. The more groups you look at, the higher the expected variation. Likewise, there's a drive on that set of charts with almost 5% AFR -- but only for 2021, versus in 2020 it had 0%.

If you play around with subgroup statistics enough, you can show that horse dewormer is a cure for covid.



You want to see a defective drives, look back at the old blogs from 2015-2017 and check the seagate failure rates.
The average AFR across their entire fleet is only 1.01%, and the Samsung EXOS drives are double that - and it's in turn also dfiferent from the median AFR, because the only drives that have a higher AFR are 14TB Seagate drives.

The point being, that Seagate are still doing worse than their competitors, and these are the drives that the youtuber was saying there was absolutely nothing wrong with, because he'd received the drives from Samsung so they were perfectly fine and definitely not worse than the competition, nope, not at all.
That is what I was laughing about.

BlankSystemDaemon fucked around with this message at 21:40 on Feb 5, 2022

Klyith
Aug 3, 2007

GBS Pledge Week

BlankSystemDaemon posted:

The average AFR across their entire fleet is only 1.01%, and the Samsung EXOS drives are double that - and it's in turn also dfiferent from the median AFR, because the only drives that have a higher AFR are 14TB Seagate drives.

The point being, that Seagate are still doing worse than their competitors, and these are the drives that the youtuber was saying there was absolutely nothing wrong with, because he'd received the drives from Samsung so they were perfectly fine and definitely not worse than the competition, nope, not at all.

Their AFR for their entire pool has ranged between 0.93% and 1.89% over the past 4 years. That particular model has been above-average every year it's been on the charts. It's not brilliant, and Seagate seems unquestionably worse than HGST for reliability -- hard to say anything about WD since Backblaze almost never uses them.

Still, 2%. And last year, with 4 times as many drive-hours, 1%.

The youtuber had a 100% failure rate, or at least a 100% errors in the raid rate (and a 0% dead-drive rate). One of these things is not like the other. I think it's crazy to blame the drives, when a much simpler explanation is that the youtuber with a detailed history of setting up braindead-stupid configurations of hardware and software hosed it up. This is the guy who made a striped raid0 of raid5s for his production server. They do idiotic stuff all the time, quite possibly on purpose for the content.

Between seagate and Linus I know which one I'd blame.

Former Human
Oct 15, 2001

Linus is definitely a dummy who thinks his staff is qualified to do projects way above their heads and then never learns his lesson.

Seagate is a crapshoot compared to the competition, though. Some of their models had a ~6% failure rate in under ten months of use in previous BackBlaze charts, which is shocking. The highest AFR for a Seagate drive in the 2021 chart is a 14TB model at 4.8%. That should be worrying to any consumer. Yeah that might not be every model but why take a chance when there are other brands?

I do wonder why BackBlaze rarely uses WD models, since they use HGST, which is owned by WD and branded as such. Does HGST still handle its own manufacturing?

mobby_6kl
Aug 9, 2009

by Fluffdaddy

Former Human posted:

Linus is definitely a dummy who thinks his staff is qualified to do projects way above their heads and then never learns his lesson.

Seagate is a crapshoot compared to the competition, though. Some of their models had a ~6% failure rate in under ten months of use in previous BackBlaze charts, which is shocking. The highest AFR for a Seagate drive in the 2021 chart is a 14TB model at 4.8%. That should be worrying to any consumer. Yeah that might not be every model but why take a chance when there are other brands?

I do wonder why BackBlaze rarely uses WD models, since they use HGST, which is owned by WD and branded as such. Does HGST still handle its own manufacturing?

I think they said before it's just supplies and/or pricing, they aren't just avoiding WD. That said I went with HGST based on the survey results just to be (a bit) safer.

Klyith
Aug 3, 2007

GBS Pledge Week

Former Human posted:

The highest AFR for a Seagate drive in the 2021 chart is a 14TB model at 4.8%.

That one is one of their low-count drives though, the confidence spread is all the way from 3-6%.

Former Human posted:

That should be worrying to any consumer. Yeah that might not be every model but why take a chance when there are other brands?

Because there's a reason the original acronym was Redundant Array of Inexpensive Disks? An intelligent storage strategy should accept the failure-prone nature of hard drives. TBQH they're all unreliable mechanical devices, the difference is just in degree.

If a seagate drive is 5% less expensive than the equivalent other brand, it's a rational choice to buy it. By the odds you come out ahead. People also buy & shuck externals that have 2 year warranty coverage, and that makes sense if it's 2/3rds or less the price. OTOH buying huge drives for your NAS where you are actually worried about reliability, because they're expensive to replace, doesn't make sense.



(I'm just gonna note, because I may sound like The Seagate Defender here: the only seagate drives I have right now are in the old sub-TB pile, and that the last ones I bought were from that ~2015 run of complete poo poo. Which died.)

Warmachine
Jan 30, 2012



What's the joke? "Hard drives are a failure device with data storage as a side effect."

Combat Pretzel
Jun 23, 2004

No, seriously... what kurds?!

Klyith posted:

This is the guy who made a striped raid0 of raid5s for his production server.
Well, when you put multiple RAID-Z vdevs into a pool, that's more or less what it also is. The only difference here is that ZFS is more resilient to one vdev failing due to multiple copies of metadata (--edit: a "traditional" filesystem just plain blows up and makes recovery a pain in the rear end, ZFS can at least still limp along).

Former Human posted:

Linus is definitely a dummy who thinks his staff is qualified to do projects way above their heads and then never learns his lesson.
Yeah, this kind of poo poo is OK at home. But in a production setup, there's a reason why there's specialized IT people for this.

SlowBloke posted:

He got a crate of exos, the equivalent of wd golds. Ironwolf are a lower tier in their catalog.
All this hard drive crap is pissing me off. I went with the 10TB ones, because that's when they switch to helium currently. I was hoping that this is just a relabelled enterprise drive of theirs, just with less warranty.

Combat Pretzel fucked around with this message at 11:45 on Feb 6, 2022

BlankSystemDaemon
Mar 13, 2009



Klyith posted:

Their AFR for their entire pool has ranged between 0.93% and 1.89% over the past 4 years. That particular model has been above-average every year it's been on the charts. It's not brilliant, and Seagate seems unquestionably worse than HGST for reliability -- hard to say anything about WD since Backblaze almost never uses them.

Still, 2%. And last year, with 4 times as many drive-hours, 1%.

The youtuber had a 100% failure rate, or at least a 100% errors in the raid rate (and a 0% dead-drive rate). One of these things is not like the other. I think it's crazy to blame the drives, when a much simpler explanation is that the youtuber with a detailed history of setting up braindead-stupid configurations of hardware and software hosed it up. This is the guy who made a striped raid0 of raid5s for his production server. They do idiotic stuff all the time, quite possibly on purpose for the content.

Between seagate and Linus I know which one I'd blame.
The youtuber had four striped 15-wide raidz3 arrays with no monitoring to give warnings (both via a bell event (can be audible via a speaker or visual via a fault LED), as well a sending email) when a drive started giving off pre-failure indicators (from either the READ/WRITE/CKSUM columns in zpool status, or from S.M.A.R.T attributes), no hot-spares configured (they should've had at least one per vdev, preferably two), was running an old version that's unlikely to include the microarchitecture-optimized RAID primitives and certainly doesn't include DRAID (which is very explicitly made for these kinds of bulk storage setups, with recordsize=10MB on the spinning rust and metadata stored on SSDs using allocation classes - all of which helps speed up resilvering).
The pool failed because one of those raidz3 vdevs had more than 3 drives fail, not because all the drives failed.

Their storage was designed to fail, even before we begin talking about the drives they were given and them insisting that the drives are fine despite the fact that there's evidence (higher than average/mean AFR) to suggest that those drives aren't fine.
Also, please re-read what BackBlaze mean by Annualized Failure Rate - it's not the same as Annual Failure Rate, and accounts for the power-on time already, so you're not supposed to factor it again.
And speaking specifically of the higher AFR for the Seagates, the article I linked explicitly talked about how the AFR compared to the Q3 numbers dropped after they updated the firmware - something the youtuber is unlikely to have done given that they didn't update the rest of the system either).

BlankSystemDaemon fucked around with this message at 13:10 on Feb 6, 2022

I would blow Dane Cook
Dec 26, 2008
ZFS is the gift that keeps on giving.

FuturePastNow
May 19, 2014


Linus just gets more content when it fails again

yummycheese
Mar 28, 2004

FuturePastNow posted:

Linus just gets more content when it fails again

yea. and it keeps the thread alive.

Ive worked with a lot of netapp equipment and at one point we had 20 standard 42u cabinets of it setup in the datacenter. Me and one other guy installed,configured,maintained the things for years with zero issues.

enterprise storage works swimmingly and frankly its so boring and unsexy that the place i worked at frequently forgot we hand this many million dollar/ multi petabyte storage array cranking along 24/7

Linus’s thing looks like rickety junk in comparison and frankly i’d wonder if its even cheaper to DIY.

tape backups and enterprise storage. two things that just work but linus hasent managed to figure out for going on a half decade

BlankSystemDaemon
Mar 13, 2009



If he wants something that's bound to fail, he should be using BTRFS.

Klyith
Aug 3, 2007

GBS Pledge Week

BlankSystemDaemon posted:

The youtuber had four striped 15-wide raidz3 arrays with no monitoring to give warnings (both via a bell event (can be audible via a speaker or visual via a fault LED), as well a sending email) when a drive started giving off pre-failure indicators (from either the READ/WRITE/CKSUM columns in zpool status, or from S.M.A.R.T attributes), no hot-spares configured (they should've had at least one per vdev, preferably two), was running an old version that's unlikely to include the microarchitecture-optimized RAID primitives and certainly doesn't include DRAID (which is very explicitly made for these kinds of bulk storage setups, with recordsize=10MB on the spinning rust and metadata stored on SSDs using allocation classes - all of which helps speed up resilvering).
The pool failed because one of those raidz3 vdevs had more than 3 drives fail, not because all the drives failed.

Ah, and that's consistent with their rebuild attempt throwing errors on every drive in the array? I don't have knowledge in this subject so I was taking Linus's word that errors distributed across all drives meant it wasn't a single drive failure.

(The RAID0 of RAID5s thing I was talking about was their much-older data disaster where they were doing something like using hardware raid controllers for the inner arrays and then unraid for the outer array. It might have been the other way around, a RAID5 of RAID0s? Whatever it was, people pointed out that it was incredibly dumb and terrible for redundancy since any 2 drives or 1 controller failure would kill it.)


BlankSystemDaemon posted:

Also, please re-read what BackBlaze mean by Annualized Failure Rate - it's not the same as Annual Failure Rate, and accounts for the power-on time already, so you're not supposed to factor it again.
Oh, I didn't mean that. Just that it was a statistically solid result since it has even more data.


BlankSystemDaemon posted:

And speaking specifically of the higher AFR for the Seagates, the article I linked explicitly talked about how the AFR compared to the Q3 numbers dropped after they updated the firmware - something the youtuber is unlikely to have done given that they didn't update the rest of the system either).

Ok that's a lol-worthy note. :v: Also explains why backblaze removed a ton of those drives from their pool in 2021, they're getting seagate to do the fix.

BlankSystemDaemon
Mar 13, 2009



Klyith posted:

Ah, and that's consistent with their rebuild attempt throwing errors on every drive in the array? I don't have knowledge in this subject so I was taking Linus's word that errors distributed across all drives meant it wasn't a single drive failure.

(The RAID0 of RAID5s thing I was talking about was their much-older data disaster where they were doing something like using hardware raid controllers for the inner arrays and then unraid for the outer array. It might have been the other way around, a RAID5 of RAID0s? Whatever it was, people pointed out that it was incredibly dumb and terrible for redundancy since any 2 drives or 1 controller failure would kill it.)

Oh, I didn't mean that. Just that it was a statistically solid result since it has even more data.

Ok that's a lol-worthy note. :v: Also explains why backblaze removed a ton of those drives from their pool in 2021, they're getting seagate to do the fix.
I'm going by what they showed in the only video I watched, which was the output of zpool status with two vdevs that had errors in the READ and WRITE columns, and one vdev that had errors in the CKSUM column.

If a device experience too high of a rate of change for either of those columns, or if the count is too high, the device is marked as as either degraded or faulted. If the device disappears completely (presumably because it catastrophically failed), or if it's taken out of rotation with zpool offline, it's marked as unavailable or offline respectively.
All states except online mean that the fault tolerance of the entire pool is compromised in some fashion.

If memory serves, their hardware RAID was a RAID5 of RAID0s (ie. RAID05 instead of RAID50, which is fairly normal although RAID60 moreso) - which itself is pretty indicative of their competence.

2% AFR is still double the AFR of the mean for their entire fleet of disks, and some of the disks have an order of magnitude lower AFR at between 0.10% and 0.20%.
While it's not the 6-10% AFR that we've seen for some manufacturers, 2% AFR is still bad.

Combat Pretzel
Jun 23, 2004

No, seriously... what kurds?!
He just published a video about using Linux' filesystem abstraction layer, that allows for just about anything to act as block storage, for swap space to enable "10TB of RAM" or some poo poo. gently caress me, they're running out of topics or something?

Adbot
ADBOT LOVES YOU

BlankSystemDaemon
Mar 13, 2009



Combat Pretzel posted:

He just published a video about using Linux' filesystem abstraction layer, that allows for just about anything to act as block storage, for swap space to enable "10TB of RAM" or some poo poo. gently caress me, they're running out of topics or something?
It's loving pathetic.

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply