Mediasonic PROBOX 8 Bay 3.5-inch Hard Drive Enclosure – USB 3.0 & eSATA Support 18TB HDD (H82-SU3S3)
Key features
- •2022 New Chipset, 2 Year Warranty.
- •Support all brand of 3.5-inch SATA hard disk drive up to 18TB per drive
- •Support all brand of 2.5-inch SATA hard drive or SSD up to 12TB per drive. (bracket adapter required sold separately)
- •Transfer rate up to 5.0Gbps via USB 3.0 and 6.0Gbps via eSATA
- •Support SATA 3 6.0Gbps hard drive transfer rate
- •One button interface selection: switch USB 3.0 or eSATA interface by pressing one button
Mediasonic PROBOX 8 Bay 3.5-inch Hard Drive Enclosure – USB 3.0 & eSATA Support 18TB HDD (H82-SU3S3)
List Price: $363.74$327.37DEALYou Save: $36.37 (10%)
Free shippingFree Returns – 30 daysFree Order CancellationSecure Payment2–3 Days DeliveryGet It June 23, 2026In Stock (1)No marketing spamNo account requiredFulfilment by FedEx / Amazon / UPS / ShipwirePayPal / Card Buyer Protection
Customer Reviews
Reviews sourced from verified Amazon purchasers4.1
out of 5
Based on 10 reviews
5★
70%
4★
30%
3★
0%
2★
0%
1★
0%
Bright Lights, Good Value,
Ambrose McNibble✓ Verified Purchase•September 30, 2023
I add my voice to the small group who think that a drive activity indicator does not need to be bright enough to read the fine print in a EULA. There is a six-inch piece of black electrical tape in the future of my ProBox.
The unit came with a map-fold sheet labeled 'Quick Installation Guide' on the front, and a 'Thank You' note. Both had different web links that offered a manual. I could not make either of these links work. When I came to write this review I discovered a bullet in the 'About This Item' section that offered a link to a manual. After scrolling up and down the page for a while I found the link rendered in the smallest type, just where the bullet said it would be. This link worked for me and led to a PDF version of the same map-fold sheet that came with the unit. In spite of what it says on the front, that sheet apparently is the manual. As such it offers pretty good graphics with mediocre and obviously translated accompanying text. The most glaring fault is the total absence of technical specifications. It does have enough information to get the unit set up and running.
One of the things you may need the manual for is how to get the unit open. It features a unique set of molded plastic latches, one for the hinged front panel (which comes off when you open it) and a different style set-of-two to remove the plate that holds the drives in their bays. These appear fairly sturdy but I wonder about their longevity in the warm environment. Fortunately they don't appear to be necessary for the operation of the unit.
As a person with experience using external drive enclosures I found the setup reasonably easy. The unit and all drives were immediately recognized by my eSATA I/O card and have run seamlessly for about a week.
I am not a fan of power bricks and this one looks particularly vexing because it uses a right-angle connector to plug into the side of the unit at the back. This means you can't put that side up against anything or have two units side-by-side without a gap.
I read some comments about overheating. I have not experienced this. I loaded the ProBox with four 7200RPM terabyte-sized drives and set it to a multi-hour backup task. I never saw temperatures in excess of 41°C. It didn't even ramp up the automatic fans. In any case there is a manual fan control that can turn them up to a dull roar.
I used eSATA to hook the ProBox to my MacPro and discovered I had to disable the automatic shut-off feature because it would not wake when the computer came out of sleep mode. This might work better with a newer computer or over USB. This isn't a big power-waster because the computer system spins down the drives in sleep mode.
I found the operation of some of the front panel push buttons less than intuitive.
To change the interface mode you must hold the button for several seconds.
To power the unit on you push and release the power button. To power it off you must push and hold the power button.
Bottom Line:
In spite of some quirky design choices, overall this appears to be a reasonably well constructed unit at an attractive price. I liked the first one well enough I ordered another.
The unit came with a map-fold sheet labeled 'Quick Installation Guide' on the front, and a 'Thank You' note. Both had different web links that offered a manual. I could not make either of these links work. When I came to write this review I discovered a bullet in the 'About This Item' section that offered a link to a manual. After scrolling up and down the page for a while I found the link rendered in the smallest type, just where the bullet said it would be. This link worked for me and led to a PDF version of the same map-fold sheet that came with the unit. In spite of what it says on the front, that sheet apparently is the manual. As such it offers pretty good graphics with mediocre and obviously translated accompanying text. The most glaring fault is the total absence of technical specifications. It does have enough information to get the unit set up and running.
One of the things you may need the manual for is how to get the unit open. It features a unique set of molded plastic latches, one for the hinged front panel (which comes off when you open it) and a different style set-of-two to remove the plate that holds the drives in their bays. These appear fairly sturdy but I wonder about their longevity in the warm environment. Fortunately they don't appear to be necessary for the operation of the unit.
As a person with experience using external drive enclosures I found the setup reasonably easy. The unit and all drives were immediately recognized by my eSATA I/O card and have run seamlessly for about a week.
I am not a fan of power bricks and this one looks particularly vexing because it uses a right-angle connector to plug into the side of the unit at the back. This means you can't put that side up against anything or have two units side-by-side without a gap.
I read some comments about overheating. I have not experienced this. I loaded the ProBox with four 7200RPM terabyte-sized drives and set it to a multi-hour backup task. I never saw temperatures in excess of 41°C. It didn't even ramp up the automatic fans. In any case there is a manual fan control that can turn them up to a dull roar.
I used eSATA to hook the ProBox to my MacPro and discovered I had to disable the automatic shut-off feature because it would not wake when the computer came out of sleep mode. This might work better with a newer computer or over USB. This isn't a big power-waster because the computer system spins down the drives in sleep mode.
I found the operation of some of the front panel push buttons less than intuitive.
To change the interface mode you must hold the button for several seconds.
To power the unit on you push and release the power button. To power it off you must push and hold the power button.
Bottom Line:
In spite of some quirky design choices, overall this appears to be a reasonably well constructed unit at an attractive price. I liked the first one well enough I ordered another.
600~800MB/s from my 4-disk RAID 0. Perfect for Mac.
David✓ Verified Purchase•September 20, 2023
My search for a RAID enclosure for my M1 iMac started a few weeks ago and ended here. Technically this is not a HW RAID but actually works better for Mac in my view. I like it so much I bought another Mediasonic Probox with USB 3.0 (HF2-SU3S3) for my backup/Time Machine drives.
Pros:
1. Fast - truly USB 3.1 gen 2 speed. With my 4 Ultrastar 7200rpm drives configured as RAID 0 by MacOS, I'm getting 500~600MB/s for writing and 700~800MB/s for reading measured by Blackmagic. This is what I had hoped since individually each would get about 230MB/s. And it's way better than 250MB/s I got from the other true HW RAID box that costed me more. Granted that box was USB 3.0 but still I had expected 500MB/s.
2. Stable. With the right MacOS RAID configuration (I use Mac OS Extended Journaled as the format and increased chunk size from default 32 to 256), it's been very stable with accumulatively >100TB transferred in the last 2 weeks.
3. Flexible. Technically this is a Mac feature and I expect all similar HDD enclosures have this benefit over HW RAID. Because each drive shows up on Mac individually, all pre-configured RAIDs work without change. No reformatting required. In future when more storage and speed are needed, you can simply add another box.
4. Good looks and instructions. The package actually looks a real product unlike some other RAID boxes I saw.
Could be improved:
1. Heat dissipation like many have mentioned especially with door closed. However I don't consider it a con and am keeping my 5-star review because 1) it's still way better than what I had before or the other HW RAID box, 2) the box offers various settings that would work in most, if not all, situations, and 3) I made a small modification that alleviated this issue somewhat (more below).
First about my setup. I keep hard drives in a closed cabinet for both maintaining a clean iMac look and keeping the noise down. Needless to say, heat dissipation is especially important to me. The cabinet has air circulation fans and I used to place the external hard drives on a 3-fan laptop cooling pad but honestly didn't think that was adequate. Those hard drives had no fans and were reliant on passive cooling. The g-drives were often hot to the touch.
This time I decided to actually figure out how hot they get. For lack of a temperature monitoring utility on Mac I used a Braun ThermoScan that seems to give accurate readings except it caps at 108F. Here is what I found:
1. With door closed, the temperature does rise quickly. With door open it works significantly better albeit the noise level goes up. Since I have it in a cabinet and noise is not issue, I'm keeping the door open.
2. The Auto Fan setting seems to work as expected. According to the spec, level 1 is 55C. In my testing it goes to level 2 when it's slightly above 108F (43C). I've never seen it go to level 3 even with door closed though. If I can trust the spec that tells me it's been able to maintain below 55C. According to Ultrastar data sheet, the operation temperature for the drive is 5~60C so that seems okay.
3. If you want to be more conservative and don't mind the noise, I find setting fan to manual level 3, I can keep the temperature below 108F even after hours of copying TBs of data. Or you can set to level 2 and it'd go beyond 108F only in extreme conditions.
4. The bottom drive was about 3F higher than the top, and I think it's because the fan is mounted high and the air tends to flow up. This is where I think it could be improved - I drilled holes on the bottom of the box and that improved the circulation quite a bit. The shell is made of metal but the bottom is conveniently of plastic. I'm sure this voided my warranty but is worth it. With this modification you can have the door closed to keep the noise down.
5. It has a power sync feature that would shut down if Mac puts the drives to sleep. That's currently my default setting and it works very well.
All in all I'm very happy with this (and the USB 3.0 model as well). I'll buy another one if I run out of space again, or upgrade to 8-bay USB 3.1 gen 2 model if they have one.
Pros:
1. Fast - truly USB 3.1 gen 2 speed. With my 4 Ultrastar 7200rpm drives configured as RAID 0 by MacOS, I'm getting 500~600MB/s for writing and 700~800MB/s for reading measured by Blackmagic. This is what I had hoped since individually each would get about 230MB/s. And it's way better than 250MB/s I got from the other true HW RAID box that costed me more. Granted that box was USB 3.0 but still I had expected 500MB/s.
2. Stable. With the right MacOS RAID configuration (I use Mac OS Extended Journaled as the format and increased chunk size from default 32 to 256), it's been very stable with accumulatively >100TB transferred in the last 2 weeks.
3. Flexible. Technically this is a Mac feature and I expect all similar HDD enclosures have this benefit over HW RAID. Because each drive shows up on Mac individually, all pre-configured RAIDs work without change. No reformatting required. In future when more storage and speed are needed, you can simply add another box.
4. Good looks and instructions. The package actually looks a real product unlike some other RAID boxes I saw.
Could be improved:
1. Heat dissipation like many have mentioned especially with door closed. However I don't consider it a con and am keeping my 5-star review because 1) it's still way better than what I had before or the other HW RAID box, 2) the box offers various settings that would work in most, if not all, situations, and 3) I made a small modification that alleviated this issue somewhat (more below).
First about my setup. I keep hard drives in a closed cabinet for both maintaining a clean iMac look and keeping the noise down. Needless to say, heat dissipation is especially important to me. The cabinet has air circulation fans and I used to place the external hard drives on a 3-fan laptop cooling pad but honestly didn't think that was adequate. Those hard drives had no fans and were reliant on passive cooling. The g-drives were often hot to the touch.
This time I decided to actually figure out how hot they get. For lack of a temperature monitoring utility on Mac I used a Braun ThermoScan that seems to give accurate readings except it caps at 108F. Here is what I found:
1. With door closed, the temperature does rise quickly. With door open it works significantly better albeit the noise level goes up. Since I have it in a cabinet and noise is not issue, I'm keeping the door open.
2. The Auto Fan setting seems to work as expected. According to the spec, level 1 is 55C. In my testing it goes to level 2 when it's slightly above 108F (43C). I've never seen it go to level 3 even with door closed though. If I can trust the spec that tells me it's been able to maintain below 55C. According to Ultrastar data sheet, the operation temperature for the drive is 5~60C so that seems okay.
3. If you want to be more conservative and don't mind the noise, I find setting fan to manual level 3, I can keep the temperature below 108F even after hours of copying TBs of data. Or you can set to level 2 and it'd go beyond 108F only in extreme conditions.
4. The bottom drive was about 3F higher than the top, and I think it's because the fan is mounted high and the air tends to flow up. This is where I think it could be improved - I drilled holes on the bottom of the box and that improved the circulation quite a bit. The shell is made of metal but the bottom is conveniently of plastic. I'm sure this voided my warranty but is worth it. With this modification you can have the door closed to keep the noise down.
5. It has a power sync feature that would shut down if Mac puts the drives to sleep. That's currently my default setting and it works very well.
All in all I'm very happy with this (and the USB 3.0 model as well). I'll buy another one if I run out of space again, or upgrade to 8-bay USB 3.1 gen 2 model if they have one.
Nice enclosure for the price, but not really recommended for hotswap with Linux+ZFS in USB3 mode - YMMV with eSATA mode
Dave Bechtel✓ Verified Purchase•September 4, 2023
( Note: this review is for the ProBox HF2-SU3S2 4-bay JBOD USB3/eSATA Mediasonic model // Non-RAID )
--I purchased this 4-bay enclosure to test Linux + ZFS drive hotswap and see if it would make a decent DAS (direct attached storage) for an old PC. In conjunction with the "StarTech.com 2 Port PCI Express SuperSpeed USB 3.0 Card Adapter with UASP - SATA PEXUSB3S24" (also available here on Amazon) this enclosure has some limitations for ZFS.
Environment: Ubuntu 14.04-64-LTS on an underclocked Dual-core (2.1GHz) PC with 4GB of RAM; CPU is not capable of hardware virtualization, but makes a decent ZFS+Samba fileserver/testbox.
Enclosure: starting with 1 drive, adding 2 more on the fly, attached with USB3 cable to USB3 PCIe card
o 1st try, 1 WD500 Blue drive, not all the way in - ' fdisk -l ' Kernel panic (drive will not slot)
Reboot
o 2nd try, tried a different drive of the same model and it slotted OK into the enclosure. Drive was detected in both cases when the USB3 cable was hot-inserted into the USB3 PCIe card (enclosure on), AND when the USB3 cable was already in and the enclosure was powered up separately.
*NOTE* - there are a couple drives that I have that would not slot into the SATA backplane of this enclosure. YMMV. I was unable to determine why.
--The good news is, you CAN insert multiple drives one at a time and they will be detected as separate drives by Linux.
--The bad news is, the enclosure DETACHES from the USB bus when you do this, so if you have any Probox-mounted drives you may be risking file/system corruption. (CAVEAT)
Test hotswap: semi-FAIL - added 3rd drive, Green, enclosure disconnected existing drives and re-added:
Jun 30 06:24:01 p2700dual1404 kernel: [ 636.100286] usb 2-1: USB disconnect, device number 2
Jun 30 06:24:20 p2700dual1404 kernel: [ 654.344522] usb 2-1: new SuperSpeed USB device number 3 using xhci_hcd
Jun 30 06:24:20 p2700dual1404 kernel: [ 654.363691] usb 2-1: New USB device found, idVendor=152d, idProduct=0567
Jun 30 06:24:20 p2700dual1404 kernel: [ 654.363697] usb 2-1: New USB device strings: Mfr=10, Product=11, SerialNumber=5
Jun 30 06:24:20 p2700dual1404 kernel: [ 654.363702] usb 2-1: Product: USB to ATA/ATAPI Bridge
Jun 30 06:24:20 p2700dual1404 kernel: [ 654.363706] usb 2-1: Manufacturer: JMicron
Jun 30 06:24:20 p2700dual1404 kernel: [ 654.363710] usb 2-1: SerialNumber: 152D00539000
Jun 30 06:24:20 p2700dual1404 kernel: [ 654.365802] usb-storage 2-1:1.0: USB Mass Storage device detected
Jun 30 06:24:20 p2700dual1404 kernel: [ 654.366189] usb-storage 2-1:1.0: Quirks match for vid 152d pid 0567: 5000000
Jun 30 06:24:20 p2700dual1404 kernel: [ 654.366227] scsi host12: usb-storage 2-1:1.0
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.364678] scsi 12:0:0:0: Direct-Access WDC WD50 00AAKX-221CA1 0125 PQ: 0 ANSI: 6
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.364907] scsi 12:0:0:1: Direct-Access WDC WD50 01AALS-00E3A0 0125 PQ: 0 ANSI: 6
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.365122] scsi 12:0:0:2: Direct-Access WDC WD50 00AADS-00L4B1 0125 PQ: 0 ANSI: 6
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.365614] sd 12:0:0:0: Attached scsi generic sg4 type 0
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.365961] sd 12:0:0:1: Attached scsi generic sg5 type 0
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.366296] sd 12:0:0:2: Attached scsi generic sg6 type 0
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.369464] sd 12:0:0:0: [sdd] 976773168 512-byte logical blocks: (500 GB/465 GiB)
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.369559] sd 12:0:0:1: [sde] 976773168 512-byte logical blocks: (500 GB/465 GiB)
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.369687] sd 12:0:0:2: [sdf] 976773168 512-byte logical blocks: (500 GB/465 GiB)
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.369857] sd 12:0:0:0: [sdd] Write Protect is off
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.369865] sd 12:0:0:0: [sdd] Mode Sense: 67 00 10 08
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.370006] sd 12:0:0:1: [sde] Write Protect is off
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.370012] sd 12:0:0:1: [sde] Mode Sense: 67 00 10 08
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.370155] sd 12:0:0:2: [sdf] Write Protect is off
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.370162] sd 12:0:0:2: [sdf] Mode Sense: 67 00 10 08
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.370330] sd 12:0:0:0: [sdd] No Caching mode page found
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.370338] sd 12:0:0:0: [sdd] Assuming drive cache: write through
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.370603] sd 12:0:0:1: [sde] No Caching mode page found
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.370608] sd 12:0:0:1: [sde] Assuming drive cache: write through
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.370624] sd 12:0:0:2: [sdf] No Caching mode page found
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.370630] sd 12:0:0:2: [sdf] Assuming drive cache: write through
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.469253] sdf: sdf1 sdf2 sdf3 sdf4
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.487575] sde: sde1 sde2 sde3 sde4
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.520648] sd 12:0:0:2: [sdf] Attached SCSI disk
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.541244] sd 12:0:0:1: [sde] Attached SCSI disk
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.562694] sdd: sdd1 sdd9
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.580201] sd 12:0:0:0: [sdd] Attached SCSI disk
# fdisk -l /dev/sdd;fdisk -l /dev/sde;fdisk -l /dev/sdf
Disk /dev/sdd: 500.1 GB, 500107862016 bytes
256 heads, 63 sectors/track, 60563 cylinders, total 976773168 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00000000
Device Boot Start End Blocks Id System
/dev/sdd1 1 976773167 488386583+ ee GPT
Disk /dev/sde: 500.1 GB, 500107862016 bytes
255 heads, 63 sectors/track, 60801 cylinders, total 976773168 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x000afbb5
Device Boot Start End Blocks Id System
/dev/sde1 2048 41945087 20971520 83 Linux
/dev/sde2 41945088 83888127 20971520 83 Linux
/dev/sde3 83888128 125831167 20971520 83 Linux
/dev/sde4 125831168 976773119 425470976 5 Extended
/dev/sde5 125833216 130027519 2097152 82 Linux swap / Solaris
/dev/sde6 130029568 197138431 33554432 83 Linux
/dev/sde7 197140480 406855679 104857600 7 HPFS/NTFS/exFAT
/dev/sde8 406857728 976773119 284957696 83 Linux
Disk /dev/sdf: 500.1 GB, 500107862016 bytes
255 heads, 63 sectors/track, 60801 cylinders, total 976773168 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x0005094a
Device Boot Start End Blocks Id System
/dev/sdf1 * 2048 52430847 26214400 83 Linux
/dev/sdf2 52430848 104859647 26214400 83 Linux
/dev/sdf3 104859648 157288447 26214400 83 Linux
/dev/sdf4 157288448 976773119 409742336 5 Extended
/dev/sdf5 157290496 161484799 2097152 82 Linux swap / Solaris
/dev/sdf6 161486848 228595711 33554432 83 Linux
/dev/sdf7 228597760 976773119 374087680 83 Linux
# ls -al /dev/disk/by-id |grep -v part
lrwxrwxrwx 1 root root 9 Jun 30 06:14 ata-SAMSUNG_HD322HJ_S17AJB0SA23730 -> ../../sda
lrwxrwxrwx 1 root root 9 Jun 30 06:14 ata-ST3320620AS_9QF4BMH8 -> ../../sdc
lrwxrwxrwx 1 root root 9 Jun 30 06:14 ata-TSSTcorp_CDDVDW_SH-S222A -> ../../sr0
lrwxrwxrwx 1 root root 9 Jun 30 06:54 ata-WDC_WD5000AAKX-221CA1_WD-WMAYUL461873 -> ../../sdg
lrwxrwxrwx 1 root root 9 Jun 30 06:14 usb-Samsung_Flash_Drive_FIT_0355715090021777-0:0 -> ../../sdb
(The following sde,sdd,sdf are the enclosure drives):
lrwxrwxrwx 1 root root 9 Jun 30 06:42 usb-WDC_WD50_00AADS-00L4B1_152D00539000-0:1 -> ../../sde
lrwxrwxrwx 1 root root 9 Jun 30 06:42 usb-WDC_WD50_00AAKX-221CA1_152D00539000-0:0 -> ../../sdd
lrwxrwxrwx 1 root root 9 Jun 30 06:42 usb-WDC_WD50_01AALS-00E3A0_152D00539000-0:2 -> ../../sdf
lrwxrwxrwx 1 root root 9 Jun 30 06:54 wwn-0x50014ee6acf2d9c8 -> ../../sdg
lrwxrwxrwx 1 root root 9 Jun 30 06:14 wwn-0x50024e920140f1b7 -> ../../sda
--Drive read speed on the WD500 Black in the enclosure is VERY good (copying to a motherboard-attached SATA drive):
# blockdev --setra 16384 /dev/sdf # read more sectors at once / I/O optimization
# time (dd if=/dev/sdf1 of=/dev/sdg1 bs=1M;sync)
20480+0 records in
20480+0 records out
21474836480 bytes (21 GB) copied, 187.885 s, 114 MB/s
real 3m12.755s
--With straight read speed, I/O is still really good; on average (monitored with iostat) I was getting ~110MB+/sec sustained from partition 1 on the WD 500 Black drive:
# time dd if=/dev/sdf1 of=/dev/null bs=1M
20480+0 records in
20480+0 records out
21474836480 bytes (21 GB) copied, 186.264 s, 115 MB/s
real 3m6.268s
--Good news: The enclosure drives also respond to " hdparm -y /dev/sdX " (instant spindown) and should also respond to " hdparm -S (sleep time) "
--The other bad news is that you will not be able to see or do SMART testing on the enclosure drives (from Linux, anyhow):
# smartctl -a /dev/sdf
smartctl 6.2 2013-07-26 r3841 [x86_64-linux-4.2.0-36-generic] (local build)
Copyright (C) 2002-13, Bruce Allen, Christian Franke, [...]
/dev/sdf: Unknown USB bridge [0x152d:0x0567 (0x205)]
Please specify device type with the -d option.
--I need to do some further testing (zfs RAID10 with some disposable 500GB drives) and will update this review as necessary, but so far pretty impressed with this device. It's working with a mix of WD Blue, Green and Black 500GB drives.
UPDATE: Further testing with ZFS on USB3
--OK, so I created a ZFS RAIDZ1 pool out of the mixed (3)xWD500 drives:
# zpool create -f -o ashift=12 -o autoexpand=on -O atime=off -O compression=lz4 zproboxRZCOMPR raidz \
usb-WDC_WD50_00AAKX-221CA1_152D00539000-0:0 usb-WDC_WD50_00AADS-00L4B1_152D00539000-0:1 usb-WDC_WD50_01AALS-00E3A0_152D00539000-0:2
# zpool status
pool: zproboxRZCOMPR
state: ONLINE
scan: none requested
config:
NAME STATE READ WRITE CKSUM
zproboxRZCOMPR ONLINE 0 0 0
raidz1-0 ONLINE 0 0 0
usb-WDC_WD50_00AAKX-221CA1_152D00539000-0:0 ONLINE 0 0 0
usb-WDC_WD50_00AADS-00L4B1_152D00539000-0:1 ONLINE 0 0 0
usb-WDC_WD50_01AALS-00E3A0_152D00539000-0:2 ONLINE 0 0 0
errors: No known data errors
--I then copied ~16GB of data to the pool and ran a test scrub to verify no data issues; I/O was really pretty good, getting ~33MB/sec sustained from each drive and the scrub finished quickly:
scrub:
Pool: zproboxRZCOMPR - scrub started: Thu Jun 30 09:36:54 CDT 2016
scan: scrub in progress since Thu Jun 30 09:36:55 2016
11.9G scanned out of 23.5G at 102M/s, 0h1m to go
0 repaired, 50.51% done
Thu Jun 30 09:38:54 CDT 2016
scan: scrub repaired 0 in 0h3m with 0 errors on Thu Jun 30 09:40:48 2016
o Scrub zproboxRZCOMPR start: Thu Jun 30 09:36:54 CDT 2016 // Completed: Thu Jun 30 09:40:51 CDT 2016
--With everything still powered on, I then removed the middle Green drive(!)
--This Failed the whole pool - after the enclosure "bounced", drives were re-added to the system as sdg,sdh:
# la /dev/disk/by-id |grep -v part
lrwxrwxrwx 1 root root 9 Jun 30 08:45 ata-SAMSUNG_HD322HJ_S17AJB0SA23730 -> ../../sde
lrwxrwxrwx 1 root root 9 Jun 30 08:45 ata-ST3320620AS_9QF4BMH8 -> ../../sdf
lrwxrwxrwx 1 root root 9 Jun 30 08:45 ata-TSSTcorp_CDDVDW_SH-S222A -> ../../sr0
lrwxrwxrwx 1 root root 9 Jun 30 08:45 usb-Samsung_Flash_Drive_FIT_0355715090021777-0:0 -> ../../sda
lrwxrwxrwx 1 root root 9 Jun 30 09:45 usb-WDC_WD50_00AAKX-221CA1_152D00539000-0:0 -> ../../sdg
lrwxrwxrwx 1 root root 9 Jun 30 09:45 usb-WDC_WD50_01AALS-00E3A0_152D00539000-0:1 -> ../../sdh
lrwxrwxrwx 1 root root 9 Jun 30 08:45 wwn-0x50024e920140f1b7 -> ../../sde
--I tried copying more data to the pool and it hung hard:
# zpool status
pool: zproboxRZCOMPR
state: ONLINE
status: One or more devices are faulted in response to IO failures.
action: Make sure the affected devices are connected, then run 'zpool clear'.
see: http://zfsonlinux.org/msg/ZFS-8000-HC
scan: scrub repaired 0 in 0h3m with 0 errors on Thu Jun 30 09:40:48 2016
config:
NAME STATE READ WRITE CKSUM
zproboxRZCOMPR ONLINE 0 439 0
raidz1-0 ONLINE 0 4 0
usb-WDC_WD50_00AAKX-221CA1_152D00539000-0:0 ONLINE 3 4 0
usb-WDC_WD50_00AADS-00L4B1_152D00539000-0:1 ONLINE 3 2 0
usb-WDC_WD50_01AALS-00E3A0_152D00539000-0:2 ONLINE 3 2 0
errors: 439 data errors, use '-v' for a list
( had to reboot here; left the green drive out to simulate failure )
pool: zproboxRZCOMPR
state: DEGRADED
status: One or more devices could not be used because the label is missing or
invalid. Sufficient replicas exist for the pool to continue
functioning in a degraded state.
action: Replace the device using 'zpool replace'.
see: http://zfsonlinux.org/msg/ZFS-8000-4J
scan: scrub repaired 0 in 0h3m with 0 errors on Thu Jun 30 09:40:48 2016
config:
NAME STATE READ WRITE CKSUM
zproboxRZCOMPR DEGRADED 0 0 0
raidz1-0 DEGRADED 0 0 0
usb-WDC_WD50_00AAKX-221CA1_152D00539000-0:0 ONLINE 0 0 0
10918954850413458013 UNAVAIL 0 0 0 was /dev/disk/by-id/usb-WDC_WD50_00AADS-00L4B1_152D00539000-0:1-part1
usb-WDC_WD50_01AALS-00E3A0_152D00539000-0:1 ONLINE 0 0 0
errors: No known data errors
( re-inserted green drive in the enclosure here - Hotswap )
( tried ZFS scrub -- hung hard )
[ 221.415966] WARNING: Pool 'zproboxRZCOMPR' has encountered an uncorrectable I/O failure and has been suspended.
# zpool scrub -s zproboxRZCOMPR
cannot cancel scrubbing zproboxRZCOMPR: pool I/O is currently suspended
( reboot again, leaving green drive in )
pool: zproboxRZCOMPR
state: ONLINE
status: One or more devices has experienced an unrecoverable error. An
attempt was made to correct the error. Applications are unaffected.
action: Determine if the device needs to be replaced, and clear the errors
using 'zpool clear' or replace the device with 'zpool replace'.
see: http://zfsonlinux.org/msg/ZFS-8000-9P
scan: resilvered 24K in 0h0m with 0 errors on Thu Jun 30 10:02:17 2016
config:
NAME STATE READ WRITE CKSUM
zproboxRZCOMPR ONLINE 0 0 0
raidz1-0 ONLINE 0 0 0
usb-WDC_WD50_00AAKX-221CA1_152D00539000-0:0 ONLINE 0 0 0
usb-WDC_WD50_00AADS-00L4B1_152D00539000-0:1 ONLINE 0 0 1
usb-WDC_WD50_01AALS-00E3A0_152D00539000-0:2 ONLINE 0 0 0
errors: No known data errors
--After rebooting with the missing drive in again, ZFS now sees the whole pool and responds OK to scrub:
Pool: zproboxRZCOMPR - scrub started: Thu Jun 30 10:05:40 CDT 2016
pool: zproboxRZCOMPR
state: ONLINE
status: One or more devices has experienced an unrecoverable error. An
attempt was made to correct the error. Applications are unaffected.
action: Determine if the device needs to be replaced, and clear the errors
using 'zpool clear' or replace the device with 'zpool replace'.
see: http://zfsonlinux.org/msg/ZFS-8000-9P
scan: scrub repaired 0 in 0h3m with 0 errors on Thu Jun 30 10:09:35 2016
config:
NAME STATE READ WRITE CKSUM
zproboxRZCOMPR ONLINE 0 0 0
raidz1-0 ONLINE 0 0 0
usb-WDC_WD50_00AAKX-221CA1_152D00539000-0:0 ONLINE 0 0 0
usb-WDC_WD50_00AADS-00L4B1_152D00539000-0:1 ONLINE 0 0 1
usb-WDC_WD50_01AALS-00E3A0_152D00539000-0:2 ONLINE 0 0 0
errors: No known data errors
o Scrub zproboxRZCOMPR start: Thu Jun 30 10:05:40 CDT 2016 // Completed: Thu Jun 30 10:09:37 CDT 2016
Conclusion: At least in USB3 mode, I would not recommend this particular enclosure for Linux+ZFS unless you are willing to put up with reboots in case of drive failure. It's probably fine to use with new drives that have been burned-in/tested as long as you have everything backed up; for all I know it may be more reliable with swapping drives under Windows.
I wish the manufacturer had some kind of update that would not cause the whole enclosure to essentially "reboot" in case a drive is removed or inserted, the whole point of having a SATA backplane is for hotswap.
If you're serious about your data, I would also recommend plugging this into a UPS instead of just a power/surge strip. Just MHO.
UPDATE: The enclosure DOES work as you would expect in eSATA mode - hotswap on a 3-drive RAIDZ1 works without unmounting/faulting the whole ZFS pool. AND you can access the SMARTCTL data on individual drives. :-) Might be a slight performance loss, but certainly merits further testing. And of course, you need to attach it to a SATA port that supports port multiplication (which my old Gigabyte motherboard does.) Very happy now.
UPDATE 2017.April: I'm very happy with this enclosure - it's still going strong and I've been using it for laptop ZFS demos in eSATA mode :)
--I purchased this 4-bay enclosure to test Linux + ZFS drive hotswap and see if it would make a decent DAS (direct attached storage) for an old PC. In conjunction with the "StarTech.com 2 Port PCI Express SuperSpeed USB 3.0 Card Adapter with UASP - SATA PEXUSB3S24" (also available here on Amazon) this enclosure has some limitations for ZFS.
Environment: Ubuntu 14.04-64-LTS on an underclocked Dual-core (2.1GHz) PC with 4GB of RAM; CPU is not capable of hardware virtualization, but makes a decent ZFS+Samba fileserver/testbox.
Enclosure: starting with 1 drive, adding 2 more on the fly, attached with USB3 cable to USB3 PCIe card
o 1st try, 1 WD500 Blue drive, not all the way in - ' fdisk -l ' Kernel panic (drive will not slot)
Reboot
o 2nd try, tried a different drive of the same model and it slotted OK into the enclosure. Drive was detected in both cases when the USB3 cable was hot-inserted into the USB3 PCIe card (enclosure on), AND when the USB3 cable was already in and the enclosure was powered up separately.
*NOTE* - there are a couple drives that I have that would not slot into the SATA backplane of this enclosure. YMMV. I was unable to determine why.
--The good news is, you CAN insert multiple drives one at a time and they will be detected as separate drives by Linux.
--The bad news is, the enclosure DETACHES from the USB bus when you do this, so if you have any Probox-mounted drives you may be risking file/system corruption. (CAVEAT)
Test hotswap: semi-FAIL - added 3rd drive, Green, enclosure disconnected existing drives and re-added:
Jun 30 06:24:01 p2700dual1404 kernel: [ 636.100286] usb 2-1: USB disconnect, device number 2
Jun 30 06:24:20 p2700dual1404 kernel: [ 654.344522] usb 2-1: new SuperSpeed USB device number 3 using xhci_hcd
Jun 30 06:24:20 p2700dual1404 kernel: [ 654.363691] usb 2-1: New USB device found, idVendor=152d, idProduct=0567
Jun 30 06:24:20 p2700dual1404 kernel: [ 654.363697] usb 2-1: New USB device strings: Mfr=10, Product=11, SerialNumber=5
Jun 30 06:24:20 p2700dual1404 kernel: [ 654.363702] usb 2-1: Product: USB to ATA/ATAPI Bridge
Jun 30 06:24:20 p2700dual1404 kernel: [ 654.363706] usb 2-1: Manufacturer: JMicron
Jun 30 06:24:20 p2700dual1404 kernel: [ 654.363710] usb 2-1: SerialNumber: 152D00539000
Jun 30 06:24:20 p2700dual1404 kernel: [ 654.365802] usb-storage 2-1:1.0: USB Mass Storage device detected
Jun 30 06:24:20 p2700dual1404 kernel: [ 654.366189] usb-storage 2-1:1.0: Quirks match for vid 152d pid 0567: 5000000
Jun 30 06:24:20 p2700dual1404 kernel: [ 654.366227] scsi host12: usb-storage 2-1:1.0
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.364678] scsi 12:0:0:0: Direct-Access WDC WD50 00AAKX-221CA1 0125 PQ: 0 ANSI: 6
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.364907] scsi 12:0:0:1: Direct-Access WDC WD50 01AALS-00E3A0 0125 PQ: 0 ANSI: 6
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.365122] scsi 12:0:0:2: Direct-Access WDC WD50 00AADS-00L4B1 0125 PQ: 0 ANSI: 6
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.365614] sd 12:0:0:0: Attached scsi generic sg4 type 0
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.365961] sd 12:0:0:1: Attached scsi generic sg5 type 0
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.366296] sd 12:0:0:2: Attached scsi generic sg6 type 0
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.369464] sd 12:0:0:0: [sdd] 976773168 512-byte logical blocks: (500 GB/465 GiB)
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.369559] sd 12:0:0:1: [sde] 976773168 512-byte logical blocks: (500 GB/465 GiB)
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.369687] sd 12:0:0:2: [sdf] 976773168 512-byte logical blocks: (500 GB/465 GiB)
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.369857] sd 12:0:0:0: [sdd] Write Protect is off
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.369865] sd 12:0:0:0: [sdd] Mode Sense: 67 00 10 08
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.370006] sd 12:0:0:1: [sde] Write Protect is off
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.370012] sd 12:0:0:1: [sde] Mode Sense: 67 00 10 08
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.370155] sd 12:0:0:2: [sdf] Write Protect is off
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.370162] sd 12:0:0:2: [sdf] Mode Sense: 67 00 10 08
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.370330] sd 12:0:0:0: [sdd] No Caching mode page found
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.370338] sd 12:0:0:0: [sdd] Assuming drive cache: write through
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.370603] sd 12:0:0:1: [sde] No Caching mode page found
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.370608] sd 12:0:0:1: [sde] Assuming drive cache: write through
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.370624] sd 12:0:0:2: [sdf] No Caching mode page found
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.370630] sd 12:0:0:2: [sdf] Assuming drive cache: write through
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.469253] sdf: sdf1 sdf2 sdf3 sdf4
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.487575] sde: sde1 sde2 sde3 sde4
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.520648] sd 12:0:0:2: [sdf] Attached SCSI disk
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.541244] sd 12:0:0:1: [sde] Attached SCSI disk
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.562694] sdd: sdd1 sdd9
Jun 30 06:24:21 p2700dual1404 kernel: [ 655.580201] sd 12:0:0:0: [sdd] Attached SCSI disk
# fdisk -l /dev/sdd;fdisk -l /dev/sde;fdisk -l /dev/sdf
Disk /dev/sdd: 500.1 GB, 500107862016 bytes
256 heads, 63 sectors/track, 60563 cylinders, total 976773168 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00000000
Device Boot Start End Blocks Id System
/dev/sdd1 1 976773167 488386583+ ee GPT
Disk /dev/sde: 500.1 GB, 500107862016 bytes
255 heads, 63 sectors/track, 60801 cylinders, total 976773168 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x000afbb5
Device Boot Start End Blocks Id System
/dev/sde1 2048 41945087 20971520 83 Linux
/dev/sde2 41945088 83888127 20971520 83 Linux
/dev/sde3 83888128 125831167 20971520 83 Linux
/dev/sde4 125831168 976773119 425470976 5 Extended
/dev/sde5 125833216 130027519 2097152 82 Linux swap / Solaris
/dev/sde6 130029568 197138431 33554432 83 Linux
/dev/sde7 197140480 406855679 104857600 7 HPFS/NTFS/exFAT
/dev/sde8 406857728 976773119 284957696 83 Linux
Disk /dev/sdf: 500.1 GB, 500107862016 bytes
255 heads, 63 sectors/track, 60801 cylinders, total 976773168 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x0005094a
Device Boot Start End Blocks Id System
/dev/sdf1 * 2048 52430847 26214400 83 Linux
/dev/sdf2 52430848 104859647 26214400 83 Linux
/dev/sdf3 104859648 157288447 26214400 83 Linux
/dev/sdf4 157288448 976773119 409742336 5 Extended
/dev/sdf5 157290496 161484799 2097152 82 Linux swap / Solaris
/dev/sdf6 161486848 228595711 33554432 83 Linux
/dev/sdf7 228597760 976773119 374087680 83 Linux
# ls -al /dev/disk/by-id |grep -v part
lrwxrwxrwx 1 root root 9 Jun 30 06:14 ata-SAMSUNG_HD322HJ_S17AJB0SA23730 -> ../../sda
lrwxrwxrwx 1 root root 9 Jun 30 06:14 ata-ST3320620AS_9QF4BMH8 -> ../../sdc
lrwxrwxrwx 1 root root 9 Jun 30 06:14 ata-TSSTcorp_CDDVDW_SH-S222A -> ../../sr0
lrwxrwxrwx 1 root root 9 Jun 30 06:54 ata-WDC_WD5000AAKX-221CA1_WD-WMAYUL461873 -> ../../sdg
lrwxrwxrwx 1 root root 9 Jun 30 06:14 usb-Samsung_Flash_Drive_FIT_0355715090021777-0:0 -> ../../sdb
(The following sde,sdd,sdf are the enclosure drives):
lrwxrwxrwx 1 root root 9 Jun 30 06:42 usb-WDC_WD50_00AADS-00L4B1_152D00539000-0:1 -> ../../sde
lrwxrwxrwx 1 root root 9 Jun 30 06:42 usb-WDC_WD50_00AAKX-221CA1_152D00539000-0:0 -> ../../sdd
lrwxrwxrwx 1 root root 9 Jun 30 06:42 usb-WDC_WD50_01AALS-00E3A0_152D00539000-0:2 -> ../../sdf
lrwxrwxrwx 1 root root 9 Jun 30 06:54 wwn-0x50014ee6acf2d9c8 -> ../../sdg
lrwxrwxrwx 1 root root 9 Jun 30 06:14 wwn-0x50024e920140f1b7 -> ../../sda
--Drive read speed on the WD500 Black in the enclosure is VERY good (copying to a motherboard-attached SATA drive):
# blockdev --setra 16384 /dev/sdf # read more sectors at once / I/O optimization
# time (dd if=/dev/sdf1 of=/dev/sdg1 bs=1M;sync)
20480+0 records in
20480+0 records out
21474836480 bytes (21 GB) copied, 187.885 s, 114 MB/s
real 3m12.755s
--With straight read speed, I/O is still really good; on average (monitored with iostat) I was getting ~110MB+/sec sustained from partition 1 on the WD 500 Black drive:
# time dd if=/dev/sdf1 of=/dev/null bs=1M
20480+0 records in
20480+0 records out
21474836480 bytes (21 GB) copied, 186.264 s, 115 MB/s
real 3m6.268s
--Good news: The enclosure drives also respond to " hdparm -y /dev/sdX " (instant spindown) and should also respond to " hdparm -S (sleep time) "
--The other bad news is that you will not be able to see or do SMART testing on the enclosure drives (from Linux, anyhow):
# smartctl -a /dev/sdf
smartctl 6.2 2013-07-26 r3841 [x86_64-linux-4.2.0-36-generic] (local build)
Copyright (C) 2002-13, Bruce Allen, Christian Franke, [...]
/dev/sdf: Unknown USB bridge [0x152d:0x0567 (0x205)]
Please specify device type with the -d option.
--I need to do some further testing (zfs RAID10 with some disposable 500GB drives) and will update this review as necessary, but so far pretty impressed with this device. It's working with a mix of WD Blue, Green and Black 500GB drives.
UPDATE: Further testing with ZFS on USB3
--OK, so I created a ZFS RAIDZ1 pool out of the mixed (3)xWD500 drives:
# zpool create -f -o ashift=12 -o autoexpand=on -O atime=off -O compression=lz4 zproboxRZCOMPR raidz \
usb-WDC_WD50_00AAKX-221CA1_152D00539000-0:0 usb-WDC_WD50_00AADS-00L4B1_152D00539000-0:1 usb-WDC_WD50_01AALS-00E3A0_152D00539000-0:2
# zpool status
pool: zproboxRZCOMPR
state: ONLINE
scan: none requested
config:
NAME STATE READ WRITE CKSUM
zproboxRZCOMPR ONLINE 0 0 0
raidz1-0 ONLINE 0 0 0
usb-WDC_WD50_00AAKX-221CA1_152D00539000-0:0 ONLINE 0 0 0
usb-WDC_WD50_00AADS-00L4B1_152D00539000-0:1 ONLINE 0 0 0
usb-WDC_WD50_01AALS-00E3A0_152D00539000-0:2 ONLINE 0 0 0
errors: No known data errors
--I then copied ~16GB of data to the pool and ran a test scrub to verify no data issues; I/O was really pretty good, getting ~33MB/sec sustained from each drive and the scrub finished quickly:
scrub:
Pool: zproboxRZCOMPR - scrub started: Thu Jun 30 09:36:54 CDT 2016
scan: scrub in progress since Thu Jun 30 09:36:55 2016
11.9G scanned out of 23.5G at 102M/s, 0h1m to go
0 repaired, 50.51% done
Thu Jun 30 09:38:54 CDT 2016
scan: scrub repaired 0 in 0h3m with 0 errors on Thu Jun 30 09:40:48 2016
o Scrub zproboxRZCOMPR start: Thu Jun 30 09:36:54 CDT 2016 // Completed: Thu Jun 30 09:40:51 CDT 2016
--With everything still powered on, I then removed the middle Green drive(!)
--This Failed the whole pool - after the enclosure "bounced", drives were re-added to the system as sdg,sdh:
# la /dev/disk/by-id |grep -v part
lrwxrwxrwx 1 root root 9 Jun 30 08:45 ata-SAMSUNG_HD322HJ_S17AJB0SA23730 -> ../../sde
lrwxrwxrwx 1 root root 9 Jun 30 08:45 ata-ST3320620AS_9QF4BMH8 -> ../../sdf
lrwxrwxrwx 1 root root 9 Jun 30 08:45 ata-TSSTcorp_CDDVDW_SH-S222A -> ../../sr0
lrwxrwxrwx 1 root root 9 Jun 30 08:45 usb-Samsung_Flash_Drive_FIT_0355715090021777-0:0 -> ../../sda
lrwxrwxrwx 1 root root 9 Jun 30 09:45 usb-WDC_WD50_00AAKX-221CA1_152D00539000-0:0 -> ../../sdg
lrwxrwxrwx 1 root root 9 Jun 30 09:45 usb-WDC_WD50_01AALS-00E3A0_152D00539000-0:1 -> ../../sdh
lrwxrwxrwx 1 root root 9 Jun 30 08:45 wwn-0x50024e920140f1b7 -> ../../sde
--I tried copying more data to the pool and it hung hard:
# zpool status
pool: zproboxRZCOMPR
state: ONLINE
status: One or more devices are faulted in response to IO failures.
action: Make sure the affected devices are connected, then run 'zpool clear'.
see: http://zfsonlinux.org/msg/ZFS-8000-HC
scan: scrub repaired 0 in 0h3m with 0 errors on Thu Jun 30 09:40:48 2016
config:
NAME STATE READ WRITE CKSUM
zproboxRZCOMPR ONLINE 0 439 0
raidz1-0 ONLINE 0 4 0
usb-WDC_WD50_00AAKX-221CA1_152D00539000-0:0 ONLINE 3 4 0
usb-WDC_WD50_00AADS-00L4B1_152D00539000-0:1 ONLINE 3 2 0
usb-WDC_WD50_01AALS-00E3A0_152D00539000-0:2 ONLINE 3 2 0
errors: 439 data errors, use '-v' for a list
( had to reboot here; left the green drive out to simulate failure )
pool: zproboxRZCOMPR
state: DEGRADED
status: One or more devices could not be used because the label is missing or
invalid. Sufficient replicas exist for the pool to continue
functioning in a degraded state.
action: Replace the device using 'zpool replace'.
see: http://zfsonlinux.org/msg/ZFS-8000-4J
scan: scrub repaired 0 in 0h3m with 0 errors on Thu Jun 30 09:40:48 2016
config:
NAME STATE READ WRITE CKSUM
zproboxRZCOMPR DEGRADED 0 0 0
raidz1-0 DEGRADED 0 0 0
usb-WDC_WD50_00AAKX-221CA1_152D00539000-0:0 ONLINE 0 0 0
10918954850413458013 UNAVAIL 0 0 0 was /dev/disk/by-id/usb-WDC_WD50_00AADS-00L4B1_152D00539000-0:1-part1
usb-WDC_WD50_01AALS-00E3A0_152D00539000-0:1 ONLINE 0 0 0
errors: No known data errors
( re-inserted green drive in the enclosure here - Hotswap )
( tried ZFS scrub -- hung hard )
[ 221.415966] WARNING: Pool 'zproboxRZCOMPR' has encountered an uncorrectable I/O failure and has been suspended.
# zpool scrub -s zproboxRZCOMPR
cannot cancel scrubbing zproboxRZCOMPR: pool I/O is currently suspended
( reboot again, leaving green drive in )
pool: zproboxRZCOMPR
state: ONLINE
status: One or more devices has experienced an unrecoverable error. An
attempt was made to correct the error. Applications are unaffected.
action: Determine if the device needs to be replaced, and clear the errors
using 'zpool clear' or replace the device with 'zpool replace'.
see: http://zfsonlinux.org/msg/ZFS-8000-9P
scan: resilvered 24K in 0h0m with 0 errors on Thu Jun 30 10:02:17 2016
config:
NAME STATE READ WRITE CKSUM
zproboxRZCOMPR ONLINE 0 0 0
raidz1-0 ONLINE 0 0 0
usb-WDC_WD50_00AAKX-221CA1_152D00539000-0:0 ONLINE 0 0 0
usb-WDC_WD50_00AADS-00L4B1_152D00539000-0:1 ONLINE 0 0 1
usb-WDC_WD50_01AALS-00E3A0_152D00539000-0:2 ONLINE 0 0 0
errors: No known data errors
--After rebooting with the missing drive in again, ZFS now sees the whole pool and responds OK to scrub:
Pool: zproboxRZCOMPR - scrub started: Thu Jun 30 10:05:40 CDT 2016
pool: zproboxRZCOMPR
state: ONLINE
status: One or more devices has experienced an unrecoverable error. An
attempt was made to correct the error. Applications are unaffected.
action: Determine if the device needs to be replaced, and clear the errors
using 'zpool clear' or replace the device with 'zpool replace'.
see: http://zfsonlinux.org/msg/ZFS-8000-9P
scan: scrub repaired 0 in 0h3m with 0 errors on Thu Jun 30 10:09:35 2016
config:
NAME STATE READ WRITE CKSUM
zproboxRZCOMPR ONLINE 0 0 0
raidz1-0 ONLINE 0 0 0
usb-WDC_WD50_00AAKX-221CA1_152D00539000-0:0 ONLINE 0 0 0
usb-WDC_WD50_00AADS-00L4B1_152D00539000-0:1 ONLINE 0 0 1
usb-WDC_WD50_01AALS-00E3A0_152D00539000-0:2 ONLINE 0 0 0
errors: No known data errors
o Scrub zproboxRZCOMPR start: Thu Jun 30 10:05:40 CDT 2016 // Completed: Thu Jun 30 10:09:37 CDT 2016
Conclusion: At least in USB3 mode, I would not recommend this particular enclosure for Linux+ZFS unless you are willing to put up with reboots in case of drive failure. It's probably fine to use with new drives that have been burned-in/tested as long as you have everything backed up; for all I know it may be more reliable with swapping drives under Windows.
I wish the manufacturer had some kind of update that would not cause the whole enclosure to essentially "reboot" in case a drive is removed or inserted, the whole point of having a SATA backplane is for hotswap.
If you're serious about your data, I would also recommend plugging this into a UPS instead of just a power/surge strip. Just MHO.
UPDATE: The enclosure DOES work as you would expect in eSATA mode - hotswap on a 3-drive RAIDZ1 works without unmounting/faulting the whole ZFS pool. AND you can access the SMARTCTL data on individual drives. :-) Might be a slight performance loss, but certainly merits further testing. And of course, you need to attach it to a SATA port that supports port multiplication (which my old Gigabyte motherboard does.) Very happy now.
UPDATE 2017.April: I'm very happy with this enclosure - it's still going strong and I've been using it for laptop ZFS demos in eSATA mode :)
I'm impressed
Matt O✓ Verified Purchase•August 12, 2023
I'm running a NAS with one 16TB drive and three 12TB drives in a hybrid RAID 5. I needed a better backup system and didn't see adding another NAS as being cost effective. I've attached this via USB 3.0 to a Windows PC. Without a raid, Windows will see the each drive in the closure as a separate drive. You cannot expand a volume in Windows across multiple USB attached drives. You can purchase software to run in Windows that will do this, but none of the free applications do this function. I've got two 6TB drives and one 12TB drive installed in the 4 bay RAID version of these Mediasonic enclosures. RAID 0 (spanning) is setup in the enclosure so Windows sees a single 24TB volume. Worth noting, RAID 0 in any device will not have drive fault tolerance (plenty of other RAID options in this enclosure). Since it's a backup to a fault tolerant NAS, I didn't need redundant redundancy ;). I've completed the backups to this enclosure which moved a bit over 14TB and it took about 30 hours (normal). When the PC is powered off, the enclosure goes into standby. As far as sleep, maybe this is what they're talking about, maybe it's not. While I'm using it as a backup in RAID 0, I could easily see using this as budget server storage in RAID 5 attached to a decently powerful PC.
As the headline states, I'm impressed. I've attached an image of my home system as a resume of sorts, so you know this isn't my first rodeo. This Mediasonic 4 bay RAID enclosure performs the functions I needed and does it very well. Most notably is how quiet the fan runs. I actually bought the Cenmate 4 Bay RAID Hard Drive Enclosure first. Hours after I finished the backup, I was hoping the fan would slow and quiet down since the drives weren't spinning. They didn't and I repacked the enclosure to return it. The fan on that box is unnerving. I'm not a fussy guy, but the fan in the Cenmate was just too much for me.
As the headline states, I'm impressed. I've attached an image of my home system as a resume of sorts, so you know this isn't my first rodeo. This Mediasonic 4 bay RAID enclosure performs the functions I needed and does it very well. Most notably is how quiet the fan runs. I actually bought the Cenmate 4 Bay RAID Hard Drive Enclosure first. Hours after I finished the backup, I was hoping the fan would slow and quiet down since the drives weren't spinning. They didn't and I repacked the enclosure to return it. The fan on that box is unnerving. I'm not a fussy guy, but the fan in the Cenmate was just too much for me.
A great 4 bay disk case for $99.
expat-mike✓ Verified Purchase•July 31, 2023
I have WD NAS 2 bay HDD for Raid 1 however that's twice the price of this 4 bay enclosure and if you want a WD or similar 4 bay diskless enclosure its even way more money. If you don't need a standalone NAS this is a great value for money product for housing bulk storage. Using free software like FreeFileSync you can make a 'clone' between two HDD's or just one and a separate USB HDD. USB3 has significantly faster file transfers than an Ethernet NAS. So, its really a matter of preferences over functionality (NAS vs this enclosure) which to chose as the capacity of this is 18TB for each HDD.
Page 1 of 2
