1

I have this problem others also had already, but none of those solutions do work for me.

Since a few days, my JBOD case holding 4 harddisks, connected to a PC via USB, disconnects and reconnects every few minutes. Previously, this did not happen at all: except for maintenance, the disks were running 7x24 since 2015.

This is a problem, because these disks (organized as a 2-ways mirror MS Storage Space) are the target/source in a producer/consumer environment. Producers and consumers expect their pathes to exist. (This should have been programmed differently, but currently, it's the way it is.) As soon as a producer or consumer does not find its files, boom.

So at first I reckoned that this is a problem caused by a hardware component. As a first try, the disk case was removed from its original network position (attached at an Intel i7 running Win8.1) and attached to another, brand new PC (Intel i7 running Win10), using a brand new USB cable and a different Ethernet cable at a different switch (originally Netgear, this one Cisco).

The problem persisted.

Copying only the relevant files to another set of 4 harddisks organized in a new storage space and using them was the next step. To no avail, the disruptions still occur.

The next step was more complicated, namely to purchase another JBOD case (had to find one first with the appropriate dimensions etc.) Anyway: problem persists.

Then I recalled MS's energy options. (I am certain no-one changed these since 2015: it would have been me to do this task, and I didn't.) On the new machine I changed the following:

[Power Options] > [Change Plan Settings] > [Change advanced power settings] > [Turn off harddisk]: Never

and

[USB settings], [USB selective suspend setting], [Setting]: disabled

Also, via [Device Manager] > [USB controllers], I located all components having a [Power Management] tab, and unchecked "Allow the computer to turn off...".

So it's not specific to

  • PC (tested on two different Intel i7)
  • switch (Netgear, Cisco)
  • network cable in between (two different ones)
  • Windows (8.1, 10)
  • JBOD case (same brand, but two different models: one RAID-capable, the other not)
  • USB cable between PC and case (original and brand new one)
  • Power management (did I overlook something? Win 8.1 is different from Win10, but it worked with those settings)

Also, it's likely (?) not a power supply issue: the individual HDs' LEDs on the disk case continue to stay on during the disks' absences; there are even some UPS to guard against outages (installed mid 2016, so their batteries are 2.5 yrs old and should still be fine). On a power loss the blinkenlights would go out instantly (tested).

Between consumers and producers are 1 to 4 switches, depending on their location. It would be quite a big thing to replace all switches and cables. (Am not sure it these could cause the described behaviour anyway.)

Any other ideas at why this changed behavior happens just so out of blue sky after years of running fine?

What could I test next?

Improve this question
5
  • If it's still the same brand it might still be the same board and for whatever reason that might be the issue. As you're saying it disconnects an reconnects every few minutes why even bother with the network if it's a local problem? I assume you mean it disconnects and reconnects the USB device within Windows?
    – Seth
    Commented Nov 8, 2018 at 10:40
  • 1
    @Seth, the two types differ in their abilities: the original one can do RAID (but was used as JBOD), the new one cannot. I was bothering with the network to rule out external influences as far as possible. And yes, the logical USB device, not the physical case: one can watch in the task manager's Performance tab the storage space disappear (however, disconnection has occured a few seconds before the actual GUI updates). After a few seconds, the USB reconnects, and accordingly the GUI updates a few seconds later. There is no power failure, there's even an UPS to guard against such.
    – Herb
    Commented Nov 8, 2018 at 10:54
  • 1
    How hot are the drives getting? If they're overheating the enclosure might shut down. Are you using the same power adapter for both, or did you switch to the new power adapter that came with the new enclosure? Is there anything that was changed around the time the problem started occurring?
    – Layne B
    Commented Nov 8, 2018 at 14:05
  • @LayneBernardo, they get not even luke warm. If anything, they are too cold. They spin slow @ 5400 rpm (WD Red). There are no external power adapters, just power cables. The involved machines were not touched for weeks. Saves run automatically. Interestingly, the last thing changing in the network was replacing in a similar but completely unrelated JBOD box the old 4x3 TB disks by new 4x10 TB disks. However, this machine is down 99% of the time, and it's 4 weeks ago, long before the problems started. (Maybe the other disks are jealous?)
    – Herb
    Commented Nov 8, 2018 at 15:42
  • Are you regularly updating your Windows operating system? Do you keep tracks of what version of USB drivers are getting installed? Try to roll back your USB drivers. Also, is it USB 3.0 or USB 2.0?
    – Ale..chenski
    Commented Nov 9, 2018 at 6:05

0

Reset to default

You must log in to answer this question.