Types of Hard drives that will Work with the V2

[PStepanas]

The V2 is rather different to previous Beyonwiz DVRs, in that it records only to an external hard drive via a USB connection (or the network, presumably -- if you can get it working reliably).

In trying to decide what kind of drive to attach to the V2 for recording, I have some questions.

1. Does the HDD need to be formatted for Linux, or can it be Windows/NTFS?

2. Is the V2 able to put the HDD into sleep mode? Does it depend on the kind of drive used?

Specifically, I have the following options:
a) standard HDD (taken out of my T3, actually), in an enclosure with no apparent power management.
b) standard SSD, in the same enclosure (this is old, and smallish, but hasn't shown any signs of imminent death)
c) WD MyPassport Ultra -- basically a fast external USB HDD
d) either of the drives from (a) or (b) but attached via a USB dock rather than an enclosure (I suspect the dock doesn't do its own power management, either) -- I'd rather keep this attached to my PC.

As you can see, my concern is that the drive is constantly going to be on, in contrast (presumably) to the internal HDD in the T3. My guess is also that only option (c) can be put into sleep mode by the V2, in which case question 1 becomes especially relevant.

So, which is my best option?

Thanks for any advice!

[exitguy]

So, which is my best option?

SSD = Quiet, Low Power draw
https://www.msy.com.au/wd-green-wds240g ... tate-drive
https://www.msy.com.au/unitek-y-1096-us ... -converter

[adoxa]

1. Does the HDD need to be formatted for Linux, or can it be Windows/NTFS?

NTFS is fine.

2. Is the V2 able to put the HDD into sleep mode? Does it depend on the kind of drive used?

Yes to both (the V2 issues a command to enter sleep mode, but the drive still has to honour it).

[PStepanas]

@exitguyaus - I love that adaptor! I wonder if my 8 year old SSD is compatible with it.

@adoxa - Is there an easy way to test if the HDD is actually going to sleep?
Obviously, I can just feel for vibration, but I'm wondering how long I need to wait since last access.

Altogether, this is great news! Thank you.

[prl]

@adoxa - Is there an easy way to test if the HDD is actually going to sleep?
Obviously, I can just feel for vibration, but I'm wondering how long I need to wait since last access.

The default is 5 minutes, but it's user-configurable stepwise from 10 seconds to 4 hours.

But if you have timeshift enabled (which it is by default), the HDD is never idle while the PVR is running normally. Timeshift is disabled in standby, so the HDD can spin down when there are no recordings being made,

[PStepanas]

Cool. Thanks.

[sonicblue]

So, which is my best option?

SSD = Quiet, Low Power draw
https://www.msy.com.au/wd-green-wds240g ... tate-drive
https://www.msy.com.au/unitek-y-1096-us ... -converter

That WD Green is rated for 80TBW so if you left it timeshifting 24/7 on a 5mbps channel it would be dead in 4.0 years.

On top of that there are times when you're recording more than 1 channel and 5mbps could double or triple.

[prl]

My understanding is that TBW (terabytes written) specs are for the warranty period. Most devices last well beyond their warranty period.

Also, while any recordings (not just more than one) will add to the writing data rate, it will typically be for much less than 27 hours/day, and so won't affect the data written nearly as much as the recordings for timeshift.

Also, that number is calculated for a device that's running 24/7. If the device is in standby, no timeshift recording is being made. Similarly if it is in shutdown while not in use (recording or being used for viewing).

I'd expect that typical use of the SSD would write far less data than assumed in the almost-worst-case example you've used.

[sonicblue]

So apparently the formula for TBW is

(SSD capacity * program/erase cycles of the NAND) / write amplification factor

But I can't see how this would be accurate as it doesn't factor in free disk space. eg. If only 1% of the drive is free you're not going to get 80 TBW from that 1%.

Suppose a beyonwiz drive is typically half full, then perhaps you would only have 40 TBW?

[prl]

Possibly, doesn't address anything that I wrote.

[sonicblue]

I wasn't trying to address what you wrote, because I thought you made reasonable points.

[peteru]

So apparently the formula for TBW is

(SSD capacity * program/erase cycles of the NAND) / write amplification factor

But I can't see how this would be accurate as it doesn't factor in free disk space. eg. If only 1% of the drive is free you're not going to get 80 TBW from that 1%.

Suppose a beyonwiz drive is typically half full, then perhaps you would only have 40 TBW?

Wear levelling is one of the contributing factors to write amplification. It ensures that the writes are spread uniformly, which can mean that data moves around. The blocks of data that are only read by the host will still be moved around the NAND blocks by the SSD to ensure that each erase block gets roughly the same amount of erase cycles.

[sonicblue]

I'm just spitballing here, but it seems to me that putting a new write into occupied cells would still require moving those occupied cells into unoccupied cells, in which case why not just put the new write into them directly and avoid the write amplification overhead.

I'm probably oversimplifying it massively, and the controller is probably doing crazy advanced things with caches, but I still think the manufacturer's TBW figure is making some assumption about how much free space is typically available.

[leeroy]

Hello all,

I am new here and considering buying beyonwiz V2, with using Portable external HD.

The following Portable external HDs are my choices. They require to draw very low power from the V2 unit via USB3 port.
"Seagate 2TB Expansion Portable Hard Drive" or "WD Elements SE 2TB USB3.0 Portable Hard Drive"
Will the above work with beyonwiz V2?

This is my crucial point of decision of purchase.

Thank you in anticipation of your response.

[IanL-S]

I have used a 4TB Seagate 2.5" USB 3 expansion drive with my T2 and it worked perfectly. I did this while overseas (about 4 weeks) and it worked without any problems. On another T2 the 4TB WD 3.5" (self-powered), the T2 decided to stop recording for unknown reason. I think the problem is related to the particular T2 rather than the HDD being used.

[leeroy]

IanL-S
You advised "4TB Seagate 2.5" USB 3 expansion drive with my T2 and it worked perfectly"

Is it connected to the USB 3 port and powered from the port and not from a different power supply?

I need the 2.5" external portable HD to be powered from the USB 3 port.

[MrQuade]

Is it connected to the USB 3 port and powered from the port and not from a different power supply?

I need the 2.5" external portable HD to be powered from the USB 3 port.

He had it connected to a T2 which doesn't have a USB3 port.

[peteru]

USB2 supplies a maximum of 500mA - generally not enough for reliable HDD operation.
USB3 supplies a maximum of 950mA - this may be enough for some low power HDDs.

[leeroy]

Thank you all for your feedbacks
by peteru
"USB3 supplies a maximum of 950mA - this may be enough for some low power HDDs".
by IanL-S
"I have used a 4TB Seagate 2.5" USB 3 expansion drive with my T2 and it worked perfectly".
by MrQuade
"He had it connected to a T2 which doesn't have a USB3 port".

Currently, I have a two 2.5" Seagate Expansion Portable Drives 2TB. They are connected to my two Media Players (Minix NEO U9-H Power Input DC 5V/3A, UGOOS AM6 PRO Power Input DC 12V/2A) via USB3 port, and have worked successfully drawing power from the players.

This would be the HD "Seagate 2TB Expansion Portable Hard Drive (2.5")", I am proposing to connect to Beyonwiz V2 USB3 port. The specs below reviewed the compatibility:
# Connect to a USB 3.0 port to take advantage of fast data transfer speeds
# Hard Drive Interface USB 3.0
# Power & Charging Interface Micro-B USB

After your feedback and researches, I am confident this HD will work with the Beyonwiz V2.

[peteru]

Bus powered HDDs connected to a USB2 port may work fine for some users for years. Until one day they don't and in the process you lose a whole bunch of data from a HDD that is almost full.

For example, I have a 1TB WD Blue drive, model WD10JPVX, in front of me right now. The label says 5V DC, 0.55A. That's just over the USB2 maximum current, even without the power required to power a SATA to USB interface. Now, that is not the typical operating current, which will be lower. So you will likely be OK until you push the hard drive hard, like when it is almost full or when it is just spinning up. Once you get to the maximum power draw, it will very much depend on the tolerances of all the components involved as to what problems, if any, you will see.

If you have the same drive hanging off a USB3 connection, you have plenty of headroom.

[leeroy]

by peteru » Sun Sep 20, 2020 13:38
"If you have the same drive hanging off a USB3 connection, you have plenty of headroom".

Thank you for your confirmation. It is greatly appreciated.

13,/10/20 Update:
I have received my beyonwiz V2 and have connected a 2.5" Seagate Expansion Portable Drives 2TB. It worked perfectly.
Thank you all who responded.

[sonicblue]

Could anyone recommend a 3.5" HDD enclosure which doesn't have undefeatable power saving where it spins down the drive after a few minutes of disk inactivity?

I tried this one from MSY and it has the issue.

https://www.msy.com.au/simplecom-se328- ... -enclosure

It causes annoyances such as having to wait for the disk to spin up when opening the movie player, or when a recording timer fires.

I am also concerned it might cause other misbehaviour as the thread waiting for disk access would get stalled for a long time, and I am not confident there aren't edge cases where this might cause spinner of death or some other unexpected behaviour.

[peteru]

If you are willing to take an unconventional approach, you could try to get a Prolific PL2773 based enclosure. I have an oldish one from Aldi. I never had any problems with it when used with my old motherboard and PCI-e based USB3 controller. When I upgraded my system and started using the enclosure with the USB3 controller on a Z490 based motherboard, I had no end of issues. Eventually I found that those problems were due to buggy firmware on Prolific PL2773 controllers. I downloaded the tools from Prolific to flash new firmware. One of the options you can tweak when flashing new firmware is the "Automatic Suspend Timer" and there is an option to completely disable it. My guess would be that most enclosures will have it enabled and set to some relatively short value out of the factory.

I can't recommend any specific model, but it may be worth opening your misbehaving enclosure to find out if it uses a PL2773. It may be possible to "fix it".

Prolific PL2773 based enclosure tweak

Screenshot_20201012_231854.png (9.11 KiB) Viewed 1832 times

[sonicblue]

Cheers Peter. I've also emailed Simplecom to see if they can provide their own flashing tool.

edit: looks like the controller for the Simplecom is JMS578 and seems there is a custom firmware that can disable the spin down feature.

[sonicblue]

No luck so far. I've followed this guide and successfully flashed the custom firmware which OP says worked for them, but the disk still spins down, and does it sooner too (3mins vs 15mins for the original fw). Also tried flashing v173.01.00.01 and found that one has the timer set to 10mins.

It seems I need to get into a Linux environment and use this tool which patches the firmware binary to a custom spin down time (0 = disabled). However the only Linux environment I have access to is the Beyonwiz V2. Should I expect the tool to run on the V2? It seems the tool was tested on ODROID (a small single board SoC running Linux) but the tool appears to be made by the OEM (JMicron) which seems to imply it's not built for ODROID but Linux in general, and should run on the V2?

Another solution might be to just write a simple plugin for the V2 which appends to a file on the disk every few minutes to reset the spin down timer, but that seems like such an inelegant solution.

[sonicblue]

Well I decided to give it a try and it seems to have worked; idling for 25 mins without a spin down.

Code: Select all

root@beyonwizv2:~/JMS578FwUpdater# ls -l
-rwxr-xr-x    1 root     root         50688 Mar  7  2019 JMS578-Hardkenel-Release-v173.01.00.02-20190306.bin
-rwxr-xr-x    1 root     root         50688 Nov  1  2017 JMS578-v0.1.0.5.bin
-rwxr-xr-x    1 root     root       4130828 Apr 19  2018 JMS578FwUpdate
-rwxr-xr-x    1 root     root        519032 Nov  1  2017 JMS578FwUpdate.v1.00
-rw-r--r--    1 root     root         50688 Dec  5  2017 JMS578_Hardkernel_v173.01.00.01.bin

root@beyonwizv2:~/JMS578FwUpdater# df
Filesystem           1K-blocks      Used Available Use% Mounted on
/dev/root              6700544    434916   5957292   7% /
devtmpfs                469348         4    469344   0% /dev
tmpfs                       64         0        64   0% /media
tmpfs                   469624       348    469276   0% /var/volatile
/dev/sdc1            976759804 825462156 151297648  85% /media/hdd
/dev/mmcblk0p3            1004         8       996   1% /boot

root@beyonwizv2:~/JMS578FwUpdater# ./JMS578FwUpdate -d /dev/sdc1 -f ./JMS578-Hardkenel-Release-v173.01.00.02-20190306.bin -b ./simplecom_orig.bin -t 0

Update Firmware file name: ./JMS578-Hardkenel-Release-v173.01.00.02-20190306.bin
Backup Firmware file name: ./simplecom_orig.bin
Auto spin-down timer: 0 min.
Backup the ROM code sucessfully.
Programming & Compare Success!!

Here is a zip file containing everything needed to easily flash the patched firmware in Windows. Just remember to tick the 'RD Version' and 'Include JM557 NVRAM' boxes, and of course check your enclosure is using the JMS578 controller.