Hey guys, I could use some help diagnosing this issue I'm seeing with my 316. I got this unit off of eBay about a week ago and I've got it all setup and updated to the latest firmware. I noticed that 2 of the drive bays were kind of finicky when trying to get them to recognize drives that were plugged into them. I had to reinsert drives multiple times so I was trying to figure out what was going on and I did a little deeper dive with smartctl. This is what I found:
/dev# smartctl -x /dev/sdb | grep SATA SATA Version is: SATA 3.0, 6.0 Gb/s (current: 1.5 Gb/s) /dev# smartctl -x /dev/sdc | grep SATA SATA Version is: SATA 3.0, 6.0 Gb/s (current: 3.0 Gb/s) /dev# smartctl -x /dev/sdd | grep SATA SATA Version is: SATA 3.0, 6.0 Gb/s (current: 3.0 Gb/s) /dev# smartctl -x /dev/sde | grep SATA SATA Version is: SATA 3.0, 6.0 Gb/s (current: 1.5 Gb/s) /dev# smartctl -x /dev/sdf | grep SATA SATA Version is: SATA 3.0, 6.0 Gb/s (current: 3.0 Gb/s) /dev# smartctl -x /dev/sdg | grep SATA SATA Version is: SATA 3.0, 6.0 Gb/s (current: 3.0 Gb/s)
The drives that are negotiating 1.5g instead of 3.0g are in the 2 slots that have been having issues detecting drives. I tried pulling all the drives out and booting the unit with fresh drives in those 2 slots. The NAS still negotiated at 1.5g in those slots. I've also tried moving the drives to other slots and in other slots they connect at 3.0g just fine. I'm thinking there is a phyiscal problem with the connector on board. I have an older Ultra6 ReadyNAS and it negotiates 3.0g on all 6 bays. Is this an issue with the NAS or is this something specific to this model?
Thanks for the troubleshooting tips. I've installed pcitools and run the commands. From the output, it looks to me like there is only 1 controller? Also, it seems odd that /dev/sda is almost always skipped but occasionally on boot I'll have an /dev/sda and no /dev/sdg, not sure what that is all about.
# lspci | grep SATA 00:1f.2 SATA controller: Intel Corporation 82801JI (ICH10 Family) SATA AHCI Controller
# udevadm info -q path -n /dev/sdb /devices/pci0000:00/0000:00:1f.2/ata1/host0/target0:0:0/0:0:0:0/block/sdb # udevadm info -q path -n /dev/sdc /devices/pci0000:00/0000:00:1f.2/ata2/host1/target1:0:0/1:0:0:0/block/sdc # udevadm info -q path -n /dev/sdd /devices/pci0000:00/0000:00:1f.2/ata3/host2/target2:0:0/2:0:0:0/block/sdd # udevadm info -q path -n /dev/sde /devices/pci0000:00/0000:00:1f.2/ata4/host3/target3:0:0/3:0:0:0/block/sde # udevadm info -q path -n /dev/sdf /devices/pci0000:00/0000:00:1f.2/ata5/host4/target4:0:0/4:0:0:0/block/sdf # udevadm info -q path -n /dev/sdg /devices/pci0000:00/0000:00:1f.2/ata6/host5/target5:0:0/5:0:0:0/block/sdg
I wonder if the ICH10 controller is just fine but the problem is the connector on the backplane is damaged somehow? I haven't really gotten in there with a flashlight to compare the connectors for physical damage vs the other ones.
Also, do you know if this model backplane has a standard SATA/SAS breakout connector on it? Could I bypass the backplane if I take it apart and hook up a regular SATA breakout cable to it? If it is just the backplane, I can probably fix that but if it is the controller, I'm kinda hosed. I don't have the skills or equipment to reball boards/replace chips like that.
The internal flash would be what's taking the other drive letter.
Not sure where the problem is, but it does sound like it could be a hardware issue of some kind. Though I don't have a 316 to check the SATA speeds negotiated.
Before considering dismantling the unit you should check the hardware warranty status. The 316 came with a 5 year limited hardware warranty. Though that has probably expired by now.
You could try removing all the disks and seeing what you can see with a flashlight without dismantling anything.
There are no SATA cable connections anywhere within the unit, so you cannot bypass the backplane. There are also no pins on the connector dedicated to 3.0G vs. 1.5G, so it's hard to imagine how a bad connection would have that particular effect and not more problems. Frequent ATA errors would be a more likely symptom. But I don't have a good suspect as to what else would be causing it, either.
I pulled out my 316, but the drives installed (Hitachi HDS722020ALA330) don't give the negotiated speed. They simply report SATA Version is: SATA 2.6, 3.0 Gb/s. If the drives were only negotiating 1.5G, I don't know if they would add that info.
If the drives were only negotiating 1.5G, I don't know if they would add that info.
My understanding is that the way this works is that when started up the disks send a bunch of ALIGNS to the host at the disk's highest supported speed. The host attempts to lock onto that clock frequency, and send back it's own ALIGNS. If the disk doesn't get a response in a set window (2048 ALIGNS), it shifts to the next slower speed and tries again. The host can restart this process by sending a reset.
So the link basically ends up running at the highest speed that the host can lock onto.
I do think it would be good to look closely at the connectors on the disks and the backplane.
Yes, but my point was that my drives are not reporting the connected speed, just their capability, so I don't know if they are connecting at 3G or not unless they only report the connected speed when it doesn't match the max capability (they are older Hitachi drives that are only SATAII).
Your handshake description is pretty much right on, but I would say " if the disk doesn't get a response it understands, then it shifts down", so noise can play a part in a bad handshake by clobbering the messages in either direction. Bad cables are a typical source of noise, so a bad or dirty connector could be as well. But that kind of noise will typically also cause other errors such as ATA errors.
I have one RN516 (which uses the same SATA backplane as the 316) for parts and already sold the backplane because another owner had an even worse issue -- a couple drives not being recognized at all. That did fix it, so there could be something wrong with the backplane itself. But his was physically damaged when the NAS took a fall with drives in it.
I tried booting the array with one disk with nothing on it and then moving the disk around. The negotiation issue stays with the individual slots. The same drive will negotiate at 3G in all but the 2 slots I've been having issues with. Some drives I put into the system power up and the blue LED on the right hand side lights up but the system doesn't recognize them. Other drives negotiate at 1.5G instead of 3G. I think this is pointing to a hardware issue with the SATA connector on the board or possibly a sketchy capacitor or something like that? I popped open the chassis and looked at everything and it all looks good. I didn't try to disassembly anything and dig deeper but it looks like there are 2 different PCIe connectors that the backplane plugs into. So no way to easily break these out.
I think I'll try to open a case with Netgear just in case the warrany is somehow still good but I'd be willing to bet it is long expired.
If anyone has any additional sugestions or things to check I'd be happy to look into this more. Thanks for the help so far.
