I have a Supermicro X10SRL-F-O running FreeNAS for almost 3 months now and lately I've been experiencing issues with the USB buses spontaneously disconnecting. This was a significant issue when I was previously booting off a mirror of thumb drives, but after I started booting off a SATA DOM, the problem persisted.
Symptoms:
When it occurs, the following shows up in messages:
Twice, the above did not happen but the ethernet changed to to 100Mbps instead of 1Gbps and stopped responding completely. The messages log only reported DNS and route failures. It notices when I disconnect and reconnect the ethernet cable, but network never responds until reboot.
System
X10SRL-F-O, with Xeon E5 1620v3 processor, 32G of ECC Crucial RAM (4x8G), and 3 IBM M1015 cards in IT firmware (P16). Internal sensors consistently report temperatures below 45°C. BIOS and Firmware are both running the latest version (1.0c and 02.18. I updated them both in the hopes of fixing the problem). USB devices have been in total: a USB hub connecting a keyboard and mouse, CyberPower UPS, and for the first two months, two Sandisk thumb drives. No other USB devices have ever been connected to this machine.
Tech Support
I contacted Supermicro to ask if my board was faulty and the response was to immediately blame FreeNAS. In my reply I told them that I know of many others who use FreeNAS on this same MB without issue. The next reply suggested that I turn of XHCI in the BIOS (note: there is no "XHCI" setting to disable, but there is a "XHCI Handoff" that can be disabled. I asked if this is the correct setting and have not yet received a response). Does this change make sense to anyone?
So, I have nearly no confidence in Supermicro resolving this issue so I thought I'd ask here. I know people on this forum have the same board so is there some setting in my BIOS that is screwing this up? Is my MB faulty and in need of replacement (it is still in warranty but I have to act quickly)? Thanks for the help.
Symptoms:
When it occurs, the following shows up in messages:
(With the last two lines repeating numerous times) Shortly after this is logged, the system reboots, and works fine for an average of a few days before it happens again.Nov 17 17:00:03 pond ugen0.2: <vendor 0x8087> at usbus0 (disconnected)
Nov 17 17:00:03 pond uhub2: at uhub0, port 1, addr 2 (disconnected)
Nov 17 17:00:03 pond ugen1.2: <vendor 0x8087> at usbus1 (disconnected)
Nov 17 17:00:03 pond uhub3: at uhub1, port 1, addr 2 (disconnected)
Nov 17 17:00:03 pond ugen0.3: <vendor 0x0557> at usbus0 (disconnected)
Nov 17 17:00:03 pond uhub4: at uhub2, port 6, addr 3 (disconnected)
Nov 17 17:00:03 pond ugen1.3: <CPS> at usbus1 (disconnected)
Nov 17 17:00:03 pond ugen1.4: <vendor 0x05e3> at usbus1 (disconnected)
Nov 17 17:00:03 pond uhub5: at uhub3, port 2, addr 4 (disconnected)
Nov 17 17:00:03 pond ugen0.4: <vendor 0x0557> at usbus0 (disconnected)
Nov 17 17:00:03 pond ukbd0: at uhub4, port 1, addr 4 (disconnected)
Nov 17 17:00:03 pond ugen1.5: <Mitsumi Electric> at usbus1 (disconnected)
Nov 17 17:00:03 pond uhub6: at uhub5, port 4, addr 5 (disconnected)
Nov 17 17:00:03 pond ums0: at uhub4, port 1, addr 4 (disconnected)
Nov 17 17:00:03 pond ugen1.6: <Logitech> at usbus1 (disconnected)
Nov 17 17:00:03 pond ums1: at uhub6, port 1, addr 6 (disconnected)
Nov 17 17:00:03 pond ugen1.7: <Mitsumi Electric> at usbus1 (disconnected)
Nov 17 17:00:03 pond ukbd1: at uhub6, port 3, addr 7 (disconnected)
Nov 17 17:00:03 pond uhid0: at uhub6, port 3, addr 7 (disconnected)
Nov 17 17:00:04 pond usbus0: port reset timeout
Nov 17 17:00:04 pond usbus1: port reset timeout
Nov 17 17:00:04 pond uhub_reattach_port: port 1 reset failed, error=USB_ERR_TIMEOUT
Nov 17 17:00:04 pond uhub_reattach_port: device problem (USB_ERR_TIMEOUT), disabling port 1
Nov 17 17:00:04 pond uhub_reattach_port: port 1 reset failed, error=USB_ERR_TIMEOUT
Nov 17 17:00:04 pond uhub_reattach_port: device problem (USB_ERR_TIMEOUT), disabling port 1
Twice, the above did not happen but the ethernet changed to to 100Mbps instead of 1Gbps and stopped responding completely. The messages log only reported DNS and route failures. It notices when I disconnect and reconnect the ethernet cable, but network never responds until reboot.
System
X10SRL-F-O, with Xeon E5 1620v3 processor, 32G of ECC Crucial RAM (4x8G), and 3 IBM M1015 cards in IT firmware (P16). Internal sensors consistently report temperatures below 45°C. BIOS and Firmware are both running the latest version (1.0c and 02.18. I updated them both in the hopes of fixing the problem). USB devices have been in total: a USB hub connecting a keyboard and mouse, CyberPower UPS, and for the first two months, two Sandisk thumb drives. No other USB devices have ever been connected to this machine.
Tech Support
I contacted Supermicro to ask if my board was faulty and the response was to immediately blame FreeNAS. In my reply I told them that I know of many others who use FreeNAS on this same MB without issue. The next reply suggested that I turn of XHCI in the BIOS (note: there is no "XHCI" setting to disable, but there is a "XHCI Handoff" that can be disabled. I asked if this is the correct setting and have not yet received a response). Does this change make sense to anyone?
So, I have nearly no confidence in Supermicro resolving this issue so I thought I'd ask here. I know people on this forum have the same board so is there some setting in my BIOS that is screwing this up? Is my MB faulty and in need of replacement (it is still in warranty but I have to act quickly)? Thanks for the help.