Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[marko]

I'm not sure, but I think these might be related to this:

Having gotten all data off my five 2Tb drives and getting them all set up in the addonics case, I opened the JMicron tool and configured my RAID5 array...

After accepting the warnings, all drives were unmounted from Windows and the JMicron thing went to work... almost as fast as the disks were unmounted, they reappeared again, as they were, ntfs volumes intact and available, the JMicron tool continued it's thing, hanging for several minutes around the 93% mark before proclaiming that the host controller was not detecting the changes and all should be rebooted. I did this, and there was no RAID volume.

I put this down to those WD Green drives and didn't try a second time.

I have 3 x Samsung 2tb drives in my addonics case, along with a couple of WD 2tb drives.
Seeing these system hangs, always coupled with disk whining noises and "the driver reported a controller error on \Device\Harddisk7\DR7." in the event logs.

Aida64 is unable to read the SMART info on one of these Samsung drives.

When I load Hard Disk Sentinel, it can read the SMART data, but rather strangely, it reports the drive as being:

Hard Disk Summary
Hard Disk Number,4
Interface,JMICRON  RAID #4/0 [14-0]
Hard Disk Model ID,SAMSUNG HD204UI
Firmware Revision,1AQ10001
Hard Disk Serial Number,S2HGJ9BB209308
Total Size,1907726 MB
Power State,Active

While the other four drives in the addonics case are all reported as being interface "SAT Standard USB/ATA" and Aida64 is able to report their SMART info.

Hard Disk Sentinel reports that the disk health is 100% excellent, and there are no SMART failures. It does report though, that:

Problems occurred between the communication of the disk and the host 52974 times. In case of a sudden crash or reboot it is recommended to try a different, short data cable (avoid round cables, use 80 wire standard cables instead). No action required.

So, why is this drive being presented as using a JMicron interface when I have uninstalled the JMicron stuff. Is that coming from the port multiplier?
Is this drive dying?
How can I be certain that this drive is what Windows refers to as "\Device\Harddisk7\DR7"?

-marko

[glynor]

The best way to check which drive it is is to check the serial numbers on the labels, since you have the serial number.

That sounds like:

1. A cabling issue.  It might just be loose or getting poor contact.
2. Drive failure.
3. Controller failure (could be software).

In roughly that order of likelihood.  Almost certainly cabling, but you can't be too sure.  Regarding the controller weirdness, I'm not sure, but HDD Sentinel might just be guessing.  What does Device Manager say?

Just to be sure I'm clear... You turned off the "hardware" RAID on the addonics add-in card, and they're just running in independent disk mode (and then you're making the RAID in software).  Is that right?

In that case, the controller would still be the JMicron controller, and it would probably still be listed as a RAID controller (because it is, despite the mode).  But it is weird that it is showing one thing for some of the drives, and another thing for others.

If you've replaced cabling, and the drive itself checks out okay (do a real scan on the disk using an Intel controller, if you can manage), and you're still having trouble... Then you might want to check for drivers for the RAID controller.  I doubt you'll be lucky enough to find much from Addonics or whomever, but... JMicron does have some generic SATA drivers available (last updated in November).  What was the make/model of the drive controller thing again?

[InflatableMouse]

Coincidently I just filed a support case with HD Sentinel. Turns out several drivers are buggy and report false SMART counter info, sometimes. My drive just dropped to 0% health due to this (hence my support ticket).

If you haven't done it yet, check the HD Sentinel site for their advised driver versions, JMicron appears several times.

So, why is this drive being presented as using a JMicron interface when I have uninstalled the JMicron stuff. Is that coming from the port multiplier?

You may have uninstalled the JMicron driver and stuff, but didn't Windows replace it with a version from Windows itself? To be sure hardware is not used, disable it in the BIOS/EUFI.

So how many controllers do you have in that system and do you know to which ports these drives are connected to? It sounds to me like that drive is connected to the wrong controller, is that possible?

Either way, sounds like a scary situation, I hope it turns out oke.

[jmone]

FYI - One of my 5-Bay cases used two controllers, and one was the JMicron so the dirves got reported differently.  It seems the JMicron controllers are very popular esp in e-sata configs.  I fond installig the JMicron dirvers was more stable.

[glynor]

So how many controllers do you have in that system and do you know to which ports these drives are connected to? It sounds to me like that drive is connected to the wrong controller, is that possible?

No.  The controllers aren't on his motherboard.  They're in a separate case.

[marko]

Almost certainly cabling, but you can't be too sure.  Regarding the controller weirdness, I'm not sure, but HDD Sentinel might just be guessing.  What does Device Manager say?

The five USB devices are in the Addonics case. I'm guessing the virtual thing is DriveBender.
In the ATA Controllers section, I'm guessing that the two Intel entries are the coloured SATA ports on my motherboard that I have lost confidence in and no longer use.

Just to be sure I'm clear... You turned off the "hardware" RAID on the addonics add-in card, and they're just running in independent disk mode (and then you're making the RAID in software).  Is that right?

The dip switches on the contoller card are all in their default JBOD positions. I have never changed them. I am not running RAID, because when I tried, using the JMicron software, it failed, and the JMicron tool relisted all five individual drives and told me there was no RAID. I did not try a second time as you had said previously that those WD green drives probably wouldn't cut it, and settled for JBOD instead.

What was the make/model of the drive controller thing again?

I'm using the Addonics HPM-XU

I'll open the case and have a look at the cables. I'll also try swapping the "308" drive to another port and see if the issue follows the drive or the port, or goes away! I'll also try reinstalling the JMicron stuff and see if that helps.

So how many controllers do you have in that system and do you know to which ports these drives are connected to? It sounds to me like that drive is connected to the wrong controller, is that possible?

Either way, sounds like a scary situation, I hope it turns out oke.

What glynor said. The five drives are in an external case, using a port multiplier and connected to the system via a single USB3 cable. I hope it turns out OK too!

It's weird you know, it's as if the drive is going to sleep. I hear this 'spin up' whine a few times when I ask for a file off it, and the application, usually MC or Lightroom, will hang for a while, then the drive springs to life and everything is peachy as long as I keep the drive busy. That's what set me off checking for SMART errors, which in turn had me chasing JMicron stuff, and I have been unable to find any power saving settings that are enabled. In my profile, Windows is set to never sleep the HDDs, and in all the USB controllers, power saving is off. However, there is no power saving tab on the properties sheet of the USB3 controller.

-marko

[MrC]

Swap two cables - if the problem follows the cable, you identified the cable.  Otherwise it's the controller or drive to focus on.

[InflatableMouse]

Ah no RAID. That simplifies things, I thought you were running a RAID there. That even made me sweat! .

Assuming you have a license for HD Sentinel I would report the weird response you're getting using the Report Menu/Send Test Report to Developer. These guys literally responded to my mail within minutes and knew immediately what was going on. I suspect HD Sentinel may not be reporting it correctly. It's a known issue and although it doesn't happen very frequently, they are optimizing internal filters to detect the sporadic incorrect values reported by controllers. If anything, it will help them optimize their product. That's not to say there's no problem with your drive, I suspect that just may be a separate issue from a failing drive or flaky cable.

With many of the USB drives I've had (including multi disk ones) sleep was a problem. Sometimes they never went to sleep and if they did it caused problems. Eventually I just decided to not use USB drives for anything other than portable use.

I find that SATA cables are worse than any other kind of cables I've worked with. The old IDE and SCSI connectors were much more secure. SATA plugs are crap IMO. So as already mentioned, check the cables, replace them or swap them and see if the problem dissapears or moves with the cable as MrC just mentioned too.

Concerning that drive, do you think you'd be able to take the drive from the case and hook it up to an internal SATA port? SMART counters should then be read correctly. It should give you a clearer picture what is going on.

[jmone]

For what it is worth (as expected), this device uses the JMicron 393 Chipset (http://www.jmicron.com/Product_JMB393.htm)

[marko]

Well, I took the easiest trouble shooting step first...

As HDS was reporting the disk as using a JMicron interface, I reinstalled the JMicron RAID configuration utility. That's been about 12 hours now, the sleepiness is gone and the Windows event logs are quiet. Bizarre.

Aida64 still cannot read the SMART data.

I got HDS on a big discount from bits du jour. I'm sure I copped a pro license on a 75% discount, and see that it's back today on 52% discount, but only for the next hour and a bit. I am absolutely sure I'm not getting the best out of it, eventually losing patience with all those tray icons it loaded, and not impressed with it replacing all my disk icons either. I tend to run it on demand now and then to make sure things are still being reported as peachy. It's telling me that the reallocated sector count on my Intel SSD is on its way up. Drives, essential as they are, can be a right royal pain in the proverbial...

I'll keep an eye on things and see how it all fares over the next few days before I start pulling the case apart and re-routing cables etc. etc.

-marko

[InflatableMouse]

Marko,

You're talking about HD Sentinel I assume?

You can turn that all off you know .. that's what I do. I got one icon in the notification tray with the highest temp number. You can configure that icons background and foreground color and its font. It mails me a report every day at 18.00 with stats and counters and it warns me whenever a counter changes or something else happens that I need to know.

One annoying thing is that on all pc's, every once in a while it seems to loose its settings. I started backing up the configuration so I can load it each time that happens, like today on my workstation.

[mykillk]

I've personally seen low quality 30 AWG SATA Cables cause weird (and hard to diagnose) problems with hard drives. If it's not 26 AWG I won't even touch it.

[mykillk]

Dupe

[marko]

Well, it's been 8 days now since I put the JMicron app back in, and the event logs have been quiet. The lagginess has also gone.

I did find one entry in event logs that proclaimed: "The JMRAIDManager service is marked as an interactive service.  However, the system is configured to not allow interactive services.  This service may not function properly." I'm not really sure what this means. I'm guessing that Windows Seven is set that way by default as I'm sure I haven't changed any setting like that, not knowingly anyway.

The number of lost communications reported by HDS has remained static at 52479. The failure of other apps to read the SMART information, and the JMicron vs ATA interface still confuses me, but I'll let it slide at the moment as everything appears to be working as it should do. Irritating though, I like things to be 'just so' with my system, and when it feels otherwise, I like to understand why. Guess I'll let this one slide for a while. I've not opened the Addonics case at all. The SATA cables in there came with the case, already plugged into the port multiplier. All I did was hook the other ends up to the drives. I've no idea what they are, they are yellow! I always thought a SATA cable was a SATA cable... never new they came in varying 'AWG's. What's an 'AWG' when it's at home?

-marko

[mykillk]

I always thought a SATA cable was a SATA cable... never new they came in varying 'AWG's. What's an 'AWG' when it's at home?

Me too! Until I built a RAID array for work years ago and every day it would fail synchronization and start rebuilding itself. I spent several days trying to figure out the issue, and almost as a last gasp measure I replaced the SATA cables and the RAID has run flawless for the last 4 years straight now!

AWG is a standard measurement of cable thickness. The higher the AWG number, the thinner the cable is. It should be printed right on the cable.

Sounds like you pretty much fixed your problem though so I doubt there's anything wrong with your cables.

[MrC]

A model of Shuttle barebones system I bought years ago had a known batch of problematic cables.  There were many forum complaints about the matter.  The problems were eliminated by replacing the cables with higher standard cables.

AWG: american wire guage (cable thickness, essentially).

[InflatableMouse]

Out of curiousity, does AWG apply to the wire carrying the actual signals, or the isolation, or both?

I had never heard of it and the SATA cables over here make no mention of it.

[mykillk]

Out of curiousity, does AWG apply to the wire carrying the actual signals, or the isolation, or both?

I had never heard of it and the SATA cables over here make no mention of it.

Just the wiring inside.

[MrC]

Monoprice (cables) typically lists cable gauge:

   http://www.monoprice.com/products/subdepartment.asp?c_id=102&cp_id=10240

but I've never noticed it on SATA cables.  I'd be more inclined to believe the connectors are the source of problems, rather than the cable gauge in SATA cables.

[InflatableMouse]

Connectors are definitely also a problem. I know because I've replaced cables more than once. Very slight pressure against the cable (not even the connector itself) and the disk dissapeared/appeared in the OS (they are hot pluggable). Often these are the cheap cables that come with a mainboard.