[H]ard Forum Storage Showoff Thread

w1retap said:
Well, I bought some drives. First I ordered (4) 20TB Seagate Exos from Newegg during their sale. Sent them all back because 3 out of 4 were DoA. Tested them all externally as well just to confirm it wasn't my backplane or something.

Fast forward 3 weeks and I hopped on the WD Red Pro sale from Western Digital's website. Tested all these, and 3 OUT OF 4 ARE DEAD. Honestly, wtf. One screeches like a dolphin, two of them just go brrr brrr brrr, and one works fine. All packaging was pristine OEM, no visible damage or prior use.

Currently trying to RMA these and the WD keeps saying they are "Out of Region" when I try to register them for RMA. Then I call their customer support and all I get is India and they keep sending me to some foreign support line that doesn't speak English. I finally got an Indian person who spoke good enough English to understand my order number and validated they are indeed under warranty. They said they'd send me an RMA via email right after the call, yet I get nothing. So sick of this.

View attachment 613533
Click to expand...
You couldn't talk to WD sales? Should be a straight swap under their return policy if they have one.
 
Spartacus09 said:
...they've made no effort to fix it.
Click to expand...
Of course not! Warranty exchanges cost money--if company can frustrate the customer and make them give up, company wins! Short term anyways unless there's competition, or if it's par for the course, suck it up customer! No one gives 2Fs about you!
 
I talked to WD Support again, and again it was India. I was able to finally get my drives to show up in the WD Support page, however the India Tech Support tells me the RMA has to be done through their site. When I go to the site, this is what I get when I try to RMA. So, I'm just in the process of filing a credit card chargeback. This is ridiculous.

cannot-RMA - Copy.PNG
 
Ho-leee-balls. This just keeps getting better. Western Digital India called me based on my support ticket I created yesterday. Just sat on the phone with them for almost 2 hours. They said they wanted to help me create the RMA. They talked me through their website where I repeated the same steps to create the RMA, and it failed again, just as shown in the screenshot above ---^ ... next they said they'd create the RMA manually since that didn't work. I read my WD order number to them, and they said they couldn't find it. Then about 20 more minutes into the call, they said they found the order. They had me verify my email address no less than 8 times, and each time they couldn't understand my perfect english using the mil-std phonetic alphabet to read my email address. After numerous tries, finally they confirmed it.

I read them all 3 of my serial numbers I was trying to RMA. Had to do this several times, and I still don't think they got it right, because the lady couldn't pronounce the english alphabet properly. Anyhow, I just said yea that's fine, I'll validate everything is correct when I get the email from you, so I'll stay on the line to validate once you send it. She said that's not possible, because they can't send the email while on the phone. I asked why that is, and she said she had to talk to her supervisor. <on hold for 10 mintues> Then it was sure, we can do that for you, it is going to be a standard RMA with processing time of 2 to 3 weeks. I'm like no, I need these sooner than that, can I do an advanced RMA? They say no, they can't do an advanced RMA because they don't do any credit card payments there. I said, please, transfer me to someone who will. They said that's not possible. So I said.. ok, I'll just do the standard RMA since you say you can't do an advanced RMA.

So, she goes off for another 5 minutes or so talking to her supervisor and says sure, yes they can do an advanced RMA. Then she said she would have to escalate the call to the support team above their level. I say ok, that's fine. Then during the transfer it just goes back to the same lady again. She says she'll have to send me an email for the advanced RMA authorization and payment. I receive it, and it is just a link to Western Digital's support website where you put in your serial numbers as shown above which doesn't work.

I asked several times if I could be transferred to a US-based office, they said it was not possible.

I feel like I could do a Louis Rossman style video on this where he talks to the State of New York.

Edit: also, I can't even do the "30-day Money Back Guarantee" since my order doesn't show up under my account, despite it being validated as the correct order number placed on their website.

Edit 2: Thought I'd call WD's Corporate Office in California at 408-801-1000. Selected the option for support on Western Digital consumer products. It just disconnects you and there is no answer from a human, routing you back to the main menu. So I go to their enterprise support line. It routes me back to the same call center in India. lmao.. you literally can't get a hold of anyone in the US at this company unless you have a specific extension and insider information.
 
Last edited:
Geez and I thought SanDisk was bad a few years ago over a ssd that died. I think WD had owned them for a year or so at that point. At least it was just an extra email explaining to them that no I couldnt run their test program because the drive was totally dead and I repeated what I'd written in the RMA request form. I think I tested that ssd in 3 different systems with 3 different sata cables, different sata ports and a HD dock (esata and usb).
 
w1retap said:
Ho-leee-balls. This just keeps getting better. Western Digital India called me based on my support ticket I created yesterday. Just sat on the phone with them for almost 2 hours. They said they wanted to help me create the RMA. They talked me through their website where I repeated the same steps to create the RMA, and it failed again, just as shown in the screenshot above ---^ ... next they said they'd create the RMA manually since that didn't work. I read my WD order number to them, and they said they couldn't find it. Then about 20 more minutes into the call, they said they found the order. They had me verify my email address no less than 8 times, and each time they couldn't understand my perfect english using the mil-std phonetic alphabet to read my email address. After numerous tries, finally they confirmed it.

I read them all 3 of my serial numbers I was trying to RMA. Had to do this several times, and I still don't think they got it right, because the lady couldn't pronounce the english alphabet properly. Anyhow, I just said yea that's fine, I'll validate everything is correct when I get the email from you, so I'll stay on the line to validate once you send it. She said that's not possible, because they can't send the email while on the phone. I asked why that is, and she said she had to talk to her supervisor. <on hold for 10 mintues> Then it was sure, we can do that for you, it is going to be a standard RMA with processing time of 2 to 3 weeks. I'm like no, I need these sooner than that, can I do an advanced RMA? They say no, they can't do an advanced RMA because they don't do any credit card payments there. I said, please, transfer me to someone who will. They said that's not possible. So I said.. ok, I'll just do the standard RMA since you say you can't do an advanced RMA.

So, she goes off for another 5 minutes or so talking to her supervisor and says sure, yes they can do an advanced RMA. Then she said she would have to escalate the call to the support team above their level. I say ok, that's fine. Then during the transfer it just goes back to the same lady again. She says she'll have to send me an email for the advanced RMA authorization and payment. I receive it, and it is just a link to Western Digital's support website where you put in your serial numbers as shown above which doesn't work.

I asked several times if I could be transferred to a US-based office, they said it was not possible.

I feel like I could do a Louis Rossman style video on this where he talks to the State of New York.

Edit: also, I can't even do the "30-day Money Back Guarantee" since my order doesn't show up under my account, despite it being validated as the correct order number placed on their website.

Edit 2: Thought I'd call WD's Corporate Office in California at 408-801-1000. Selected the option for support on Western Digital consumer products. It just disconnects you and there is no answer from a human, routing you back to the main menu. So I go to their enterprise support line. It routes me back to the same call center in India. lmao.. you literally can't get a hold of anyone in the US at this company unless you have a specific extension and insider information.
Click to expand...
This is the nightmare that drive warranties have become. I had some trouble with them a few years ago for an HGST that was now warrant'd by WD, but they got it fairly easily compared to the hell they're putting you through. But compared to WD, I've heard Toshiba enterprise is so bad that most people don't even try. :eek:
 
Latest in the saga. Trying the email route. They don't seem to understand I purchased these drives from their webstore. I've made it clear in numerous emails and even attached my invoice. Insanity. They seem to think I'm in some bizarre country and bought them from a Jakarta market.

Screenshot_20231119-173904.png
 
w1retap said:
Latest in the saga. Trying the email route. They don't seem to understand I purchased these drives from their webstore. I've made it clear in numerous emails and even attached my invoice. Insanity. They seem to think I'm in some bizarre country and bought them from a Jakarta market.
Click to expand...
Absolutely insane--this is what any support in the third world looks like, which is none. Sad to see that anyone in the west is subjected to this garbage.

Here's a route that seems to work for a lot of companies, but really shouldn't (my wife would kill me if I told you this trick because she deals with it as head of support for a company)--find some of the principals of WD on linked in and message them. Chief of Support, Director of Support, VP of support, etc. Hit them all--you might be surprised how fast this is resolved. ;)
 
w1retap said:
I'll have to reactivate my LinkedIn. I can use leverage that I'll recommend for all multi-million dollar computer projects I manage at the Fortune 500 company I work for, we'll no longer be utilizing WD as a partner haha.
Click to expand...
That would certainly do it!! :D
 
Return the drives to the please of purchase sir! To the please of purchase!

Please of Purchase to the courtesy phone..please of purchase to the courtesy phone, please, for your token of apologies.
 
Update.. I was able to get WD email support (after about 12 back and forth emails) to give me a return w/ refund. These will be yeeted into the UPS drop box as soon as possible this morning. Holy crap what a nightmare. There's another guy on a private tracker forum I'm on dealing with the same thing right now too, and his show up as "Out of Region" upon registration as well.

Do not purchase from WD.
 
I grabbed a 16TB WD drive to shuck from Amazon during Black Friday. One of the disks in my NAS had failed, figured might as well upgrade. <12 hours between ordering and having it on my doorstep. ...Yeah. Before I even got the package opened, another disk had been marked degraded and one file had been corrupted. Rats. And of course, the 16TB deal is gone from Amazon now. Still available on WD's site, but after w1retap's issues, I ain't goin' near it. Looks like I can pick up 14TB drives from BestBuy for the same price, and they'd match everything else in my NAS except the one new 16TB drive, so I did that x4. (Had to make two orders because of their "3 per order" limit.)

Screenshot 2023-11-25 152333.png


I'm getting sick of Seagate drives dying on me. First the Constellation series, then the Barracudas, now the Ironwolfs. You'd think I'd learn my lesson.
Code: 
[root@nas Downloads]# zpool status
  pool: backup
 state: ONLINE
  scan: scrub repaired 0B in 0 days 19:36:54 with 0 errors on Sat Nov 13 23:55:08 2021
config:

        NAME               STATE     READ WRITE CKSUM
        backup             ONLINE       0     0     0
          mirror-0         ONLINE       0     0     0
            14TB-9MH4K35K  ONLINE       0     0     0
            14TB-9MH5P6BK  ONLINE       0     0     0
          mirror-1         ONLINE       0     0     0
            14TB-Y6GBZW5D  ONLINE       0     0     0
            14TB-Y6GE8DPD  ONLINE       0     0     0
          mirror-2         ONLINE       0     0     0
            14TB-Y6GK8E2C  ONLINE       0     0     0
            14TB-Y6GKB43C  ONLINE       0     0     0
          mirror-3         ONLINE       0     0     0
            14TB-Y6GL164D  ONLINE       0     0     0
            14TB-Y6GTHUBC  ONLINE       0     0     0
          mirror-4         ONLINE       0     0     0
            14TB-Y6GTNZVC  ONLINE       0     0     0
            14TB-Y6GTU4EC  ONLINE       0     0     0
          mirror-5         ONLINE       0     0     0
            14TB-Y6GTUL9C  ONLINE       0     0     0
            14TB-Y6GU0N9C  ONLINE       0     0     0

errors: No known data errors

  pool: pool
 state: DEGRADED
status: One or more devices is currently being resilvered.  The pool will
        continue to function, possibly in a degraded state.
action: Wait for the resilver to complete.
  scan: resilver in progress since Sat Nov 25 15:13:19 2023
        345G scanned at 115G/s, 157M issued at 52.3M/s, 52.2T total
        0B resilvered, 0.00% done, 12 days 02:51:03 to go
config:

        NAME                        STATE     READ WRITE CKSUM
        pool                        DEGRADED     0     0     0
          raidz1-0                  DEGRADED     0     0     0
            12TB-ZJV0VHSS           ONLINE       0     0     0
            12TB-ZCH0HBK0           ONLINE       0     0     0
            12TB-ZJV2J70M           ONLINE       0     0     0
            12TB-ZJV0VZ0Z           DEGRADED     0     0     0  too many errors
            14TB-2BGTL7WR           ONLINE       0     0     0
            replacing-5             DEGRADED     0     0     0
              12857538123529081610  UNAVAIL      0     0     0  was /dev/disk/by-vdev/12TB-ZJV0V6RQ-part1
              16TB-4JG0ZJGW         ONLINE       0     0     0

errors: No known data errors
[root@nas Downloads]# df -h
Filesystem                      Size  Used Avail Use% Mounted on
***
backup                           73T   42T   31T  58% /backup
pool                             51T   42T  8.6T  83% /pool
 
New Pool with Segate EXOS 14 12TB refurb drives from ebay. went Raid 2z. I am impressed with the speed. via SMB using a single USB 3.0 drive I am to pull 230MB/s from my Windows VM with the drive passed directly to it coping the data to VM based FreeNas Core on the same host.
 

Attachments

  • New ZFS array.png
    New ZFS array.png
    178.7 KB · Views: 1
  • ZFS Pool 2.png
    ZFS Pool 2.png
    141.7 KB · Views: 1
  • ZFS Server.png
    ZFS Server.png
    31.9 KB · Views: 1
TeeJayHoward said:
I'm getting sick of Seagate drives dying on me. First the Constellation series, then the Barracudas, now the Ironwolfs. You'd think I'd learn my lesson.
Click to expand...

I've been running Seagate enterprise drives for years now without a single failure.

First 12x ST10000NM0086's (10TB Exos X10) from ~2017 until ~2022

Then 12x ST16000NM000J (16TB Exos X18) from ~2022 until today.

That's 84 drive-years.

Maybe I've just been lucky, or maybe the key is to run Enterprise drives instead of consumer drives?
 
Zarathustra[H] said:
...the key is to run Enterprise drives instead of consumer drives?
Click to expand...
This is the key to reduced drive failures. All enterprise drives are a better spec. Want to turn that up a notch? Go SAS. Once you go SAS you'll never go back...
 
Zarathustra[H] said:
maybe the key is to run Enterprise drives instead of consumer drives?
Click to expand...
SamirD said:
This is the key to reduced drive failures. All enterprise drives are a better spec. Want to turn that up a notch? Go SAS. Once you go SAS you'll never go back...
Click to expand...
The Constellations were SAS Enterprise drives. They had a higher failure rate than the shucked WD consumer SATA drives in the same chassis.
 
New Plex Server -
1709560112184.png


Spec's
Gigabyte AX570-xtreme
AMD 3950x
32gb TridentZ
MSI 1660ti (transcoding)
2x 1tb Sabrent Rocket's
16x Seagate X22 20tb's
LSI 9340i to the BPN-SAS2 Backplane

I'm actually having difficulties with the backplane. Port 0 is super flaky on it and fails to detect the drive every 3rd boot or so. If I pop the drive out and put it back in it shows back up. Any ideas?


Initial plans were to try out NAP-IT and do ZFS on windows, but my initial testing of it was godawful slow. Just throwing up a Storage Spaces pool for comparison purposes was about 4 times faster. So I might blow it away and put Linux on it instead. I've heard the aquantia nic's kinda suck for Linux though, so I might need to find a way to cram a Melanox X-Connect or a Intel 710 in there for 10gb.
 
Phelptwan said:
I'm actually having difficulties with the backplane. Port 0 is super flaky on it and fails to detect the drive every 3rd boot or so. If I pop the drive out and put it back in it shows back up. Any ideas?
Click to expand...
Have you tried a different drive? If it's still doing that, I'd power it all down and examine that particular backplane connector for issues. If you can't spot anything I would give it a good hit with some CRC contact cleaner per the bottle's directions.
 
Phelptwan said:
New Plex Server -
View attachment 639413

Spec's
Gigabyte AX570-xtreme
AMD 3950x
32gb TridentZ
MSI 1660ti (transcoding)
2x 1tb Sabrent Rocket's
16x Seagate X22 20tb's
LSI 9340i to the BPN-SAS2 Backplane

I'm actually having difficulties with the backplane. Port 0 is super flaky on it and fails to detect the drive every 3rd boot or so. If I pop the drive out and put it back in it shows back up. Any ideas?


Initial plans were to try out NAP-IT and do ZFS on windows, but my initial testing of it was godawful slow. Just throwing up a Storage Spaces pool for comparison purposes was about 4 times faster. So I might blow it away and put Linux on it instead. I've heard the aquantia nic's kinda suck for Linux though, so I might need to find a way to cram a Melanox X-Connect or a Intel 710 in there for 10gb.
Click to expand...
Make sure the backplane is screwed in correctly(if its miss aligned could cause the drive not to make good contact) Check power plugs on the backplane. I have the same problem on my backplane. 2 ports are dead... it drives me nuts but I just live with it. Also that 1660 seems overkill. You can get used Quadro P400 2GB PCIe X16 cards for 40$ on ebay. Less power and perfect for transcoding

Also go Proxmox. then do a freenas core and pass the HBA directly to the VM. ZFS will work perfectly as long as HBA card is in IT mode (aka non raid mode)
 
Vengance_01 said:
Make sure the backplane is screwed in correctly(if its miss aligned could cause the drive not to make good contact) Check power plugs on the backplane. I have the same problem on my backplane. 2 ports are dead... it drives me nuts but I just live with it.
Click to expand...
Interesting that you mention this. I have an HP DL380 G5 that has one port that seems to be flakey but only started after it was racked and it was a bit tight--you think it could be a similar issue where the backplane just move/shifted? I haven't attempted to unrack it since I stacked other servers on top of it.
 
Vengance_01 said:
Make sure the backplane is screwed in correctly(if its miss aligned could cause the drive not to make good contact) Check power plugs on the backplane. I have the same problem on my backplane. 2 ports are dead... it drives me nuts but I just live with it. Also that 1660 seems overkill. You can get used Quadro P400 2GB PCIe X16 cards for 40$ on ebay. Less power and perfect for transcoding

Also go Proxmox. then do a freenas core and pass the HBA directly to the VM. ZFS will work perfectly as long as HBA card is in IT mode (aka non raid mode)
Click to expand...
It's the only port with issues. Tested with a range of different drives. Might just need to live with it.

1660 was free. Actually everything but the drives was free or from prior builds that were just laying around.
 
Phelptwan said:
It's the only port with issues. Tested with a range of different drives. Might just need to live with it.
Click to expand...
I'd definitely try the CRC contact cleaner. Even if all you can do is power it off and disconnect all power to it (and let everything drain out of power) and then just spray the heck out of it and let it dry overnight, it definitely won't hurt. I once had a switch contact board for one of our gas pumps that was dead and before replacing the part I tried CRC and sure enough it started working again and kept working. I ended up with just having a spare contact board. :)
 
Vengance_01 said:
Make sure the backplane is screwed in correctly(if its miss aligned could cause the drive not to make good contact) Check power plugs on the backplane. I have the same problem on my backplane. 2 ports are dead... it drives me nuts but I just live with it. Also that 1660 seems overkill. You can get used Quadro P400 2GB PCIe X16 cards for 40$ on ebay. Less power and perfect for transcoding

Also go Proxmox. then do a freenas core and pass the HBA directly to the VM. ZFS will work perfectly as long as HBA card is in IT mode (aka non raid mode)
Click to expand...

First, that is way more official looking than my media server (emby) in an old antec case.

I use truenas scale. I messed with core, but settled on scale because it handled the gpu transcode configuration easier. Then, if you do the driver unlock for more streams, gotta use scale. And transcode overkill is the name of the game...I use a 1080ti.
 
SamirD said:
I'd definitely try the CRC contact cleaner. Even if all you can do is power it off and disconnect all power to it (and let everything drain out of power) and then just spray the heck out of it and let it dry overnight, it definitely won't hurt. I once had a switch contact board for one of our gas pumps that was dead and before replacing the part I tried CRC and sure enough it started working again and kept working. I ended up with just having a spare contact board. :)
Click to expand...
Just spray it right at the faulty port?
 
Had good luck with Toshiba N series (NAS) drives. Hundreds of drives, some still in service over 5 years, even better was the HGST drives, a few of the 4TB and 5TB models over 12 years in service with zero bad blocks! I agree, Seagate is much worse in NAS environments. Ironwolf series has over 40% failure rate. They are good for RMA but warranties do run out. Thank goodness for RAID6. Cannot wait for flash to replace this spinning rust once and for all!

Also, those 20TB Exos drives should have 512MB cache!
 
Last edited:
cpufrost said:
Had good luck with Toshiba N series (NAS) drives. Hundreds of drives, some still in service over 5 years, even better was the HGST drives, a few of the 4TB and 5TB models over 12 years in service with zero bad blocks! I agree, Seagate is much worse in NAS environments. Ironwolf series has over 40% failure rate. They are good for RMA but warranties do run out. Thank goodness for RAID6. Cannot wait for flash to replace this spinning rust once and for all!

Also, those 20TB Exos drives should have 512MB cache!
Click to expand...
Imo, anything labelled 'NAS' is generally not a great drive. Better than desktop, but since these originated as desktop drives with a different firmware, their roots were never the same as true enterprise drives. Toshibas today are some HGST since they bought out part of HGST and some other part is WD, but HGST was the king. You can even tell these drives from their 3 screws on the cover, and it's the same design that's found on today's enterprise HGST/WD and even the higher end 'NAS Pro' line.

The Seagate Exos line is the same enterprise stuff that's from any other maker ime. I have 3 (5?) 16TB exos I got when the series was first introduced and they're just the same as my Toshibas/WD/HGSTs. I think the environment makes the critical difference for each brand whereas I'm sure each has a sweet spot. That being said, most oem server sas drives are Seagate so they definitely do make solid drives, but probably not as concentrated as someone like WD who is focused on consumer not enterprise business.

As far as the cache, it depends on what series. Generally, the newer drives have larger cache.
 
Yeah the NAS and X series are mechanically the same, the latter has its firmware tuned for better desktop performance. (supposedly)
Which of course, today, is obsolete. ;-)

Speaking of "enterprise" drives, the original IBM GXP series were called that by geeks/nerds because the underside sort of looks like "The Enterprise" aka NCC-1701. :-D

I remember in late 1998 it was announced that WD was coming out with an "Expert" line of enthusiast drives in 7200 RPM flavor. Since they didn't have the tech, these were rebadged Deskstar drives. I think I still have one with 14.4GB capacity. It was quite snappy AND noisy like in the same class of the Barracudas. They also released a 27.3GB version later in 1999. Of course when (IBM) dropped the infamous 75GXP series the rest was history. I personally liked those drives and did experience a few failures but the tech at the time was quite amazing.
 
cpufrost said:
Yeah the NAS and X series are mechanically the same, the latter has its firmware tuned for better desktop performance. (supposedly)
Which of course, today, is obsolete. ;-)

Speaking of "enterprise" drives, the original IBM GXP series were called that by geeks/nerds because the underside sort of looks like "The Enterprise" aka NCC-1701. :-D

I remember in late 1998 it was announced that WD was coming out with an "Expert" line of enthusiast drives in 7200 RPM flavor. Since they didn't have the tech, these were rebadged Deskstar drives. I think I still have one with 14.4GB capacity. It was quite snappy AND noisy like in the same class of the Barracudas. They also released a 27.3GB version later in 1999. Of course when (IBM) dropped the infamous 75GXP series the rest was history. I personally liked those drives and did experience a few failures but the tech at the time was quite amazing.
Click to expand...
I do not believe NAS and enterprise drives are mechanically the same. The noise level gives away the beefier enterprise components while being 'quiet' (like desktop drives), is a hallmark of NAS drives and probably a characteristic of their frailty.

Hard drives never had their noise levels as being a design consideration until the consumer came along. Before that, drives were what they were. But today 'quiet' is a design spec and that creates compromise in cost, performance, or reliability or all 3. Hence why noise isn't a design spec on sas drives.
 
Swapping from spinning rust to flash for the home NAS setup. While I wanted to build an EPYC/Threadripper setup for this, the budget simply didn't allow for it. So, I'm reusing an old workstation I had laying around instead. I figure the Xeon W-3223 in it has 64 lanes. They're 3.0, but that's fine. I'm limited by the NIC anyway. No Bifurcation, which means expensive cards, but still doable.

8x Crucial P3+ 4TB M.2 NVMe drives per Sonnet 8x4 card.
IMG_4844.JPG


Heatsink attached:
IMG_4845.JPG


Do it again so there's two maxed out cards:
IMG_4846.JPG


Toss 'em in the MacPro:
IMG_4848.JPG


And put the case back on:
IMG_4849.JPG


Then clean up the mess left behind.
IMG_4850.JPG


Verify they're all there.
ready for zfs.png


Now, how to set it up? RaidZ?
raidz-writes-at-750MBps-reads-at-300MBps.png


Eek. That's quite the performance hit. Not to mention my data hoarding would necessitate the addition of another loaded card. Hows about simple RAID-0 and I keep the spinning rust for backup?
raid0-256-writes-at-1GBps-reads-at-275MBps.png


Maybe just a plain ol' JBOD?
jbod-writes-at-1GBps-reads-at-750MBps.png


The numbers are... Well, about what I'd expect for a single NVMe drive. But the network's only 10GbE so they should be just fine. Power draw at the wall is a little under 300W. More than I wanted, but I imagine much less than the 24-disk E3v2 I was using previously. I'd grab numbers to compare but the old and new NAS are syncing this week.

Now for the next issue - why does MacOS suck so badly at SMB? iperf3 shows line speed. I can write data to RAID-0 or JBOD at over 1000MB/s. RaidZ is about where I'd expect at 750MB/s...ish. But reading from it caps out at about 750MB/s for JBOD. (And on the raid setups, it's even worse - 300MB/s!) Most of my files are 8-50GB in size. I tried lots of different block sizes, it didn't make a difference. Copying from one NVMe drive to another goes quick as. It's just over the network that things get terrible. So maybe I'll end up building that Threadripper/EPYC box eventually anyway. Shame there's not a <100W TDP chip out there with >80 PCIe 4.0+ lanes (and the associated mobo with 5+ x16 slots). If I do end up building the new rig, it's gonna be mostly wasted. I don't need cores, damnit, I need PCIe lanes!

edit: Giving SoftRAID a try. I do NOT like the idea of a yearly subscription just to be able to create raid volumes, but if the performance is there...

Screenshot 2024-12-01 at 9.29.57 PM.png
Screenshot 2024-12-01 213216.png


edit2: It isn't. That is SO much worse than even RAIDZ. And people pay money for this!?

edit3: Went back to ZFS. Increasing ashift values gave me significant performance improvements. ashift=15 was twice what ashift=13 was... But it was also half my capacity gone. Oddly, moving down to ashift=9 resulted in a performance improvement AND increased capacity. 54TiB still ain't enough, but I guess I can get rid of some of the older junk I've got in storage.
 
Last edited:
EnderW said:
Did you really pay $1600 for those cards?
Click to expand...
Yup. No bifurcation means the cheap cards are a no go. 64 total PCIe lanes means I can't get the smaller 4xM.2 cards if I want any sort of capacity. By going with the 8x cards, I can have three plus a GPU installed, so I've got some room for upgrades. By going with the Sonnet (or OWC) cards, I know they'll work with MacOS (which, admittedly, isn't a huge deal since I could just throw some flavor of *NIX on instead)
 
You must log in or register to reply here.
Back
Top