[solved] e1000e: probe of 0000:04:00.0 failed with error -2
We recently had to solve a very strange issue with a server which has been running just fine for years but after a reboot had a non working network card.
The server is SuperServer 6015B-T with a X7DBR-3 Motherboard~#
dmidecode -t 2
# dmidecode 2.12
SMBIOS 2.4 present.
Handle 0x0002, DMI type 2, 8 bytes
Base Board Information
Manufacturer: Supermicro
Product Name: X7DBR-3
Version: PCB Version
Serial Number: 0123456789
Network Controllers is :
Intel® (ESB2/Gilgal) 82563EB Dual-Port Gigabit Ethernet Controller
Supports 10BASE-T, 100BASE-TX, and 1000BASE-T, RJ45 output
Intel® I/OAT support for fast, scaleable, and reliable networking
# lspci |grep -i eth
04:00.0 Ethernet controller: Intel Corporation 80003ES2LAN Gigabit Ethernet Controller (Copper) (rev 01)
04:00.1 Ethernet controller: Intel Corporation 80003ES2LAN Gigabit Ethernet Controller (Copper) (rev 01)
Logs (from /var/log/messages or dmesg |grep -e eth -e e1000 ) are the following:
e1000e: Intel(R) PRO/1000 Network Driver - 2.3.2-k
e1000e: Copyright(c) 1999 - 2014 Intel Corporation.
Interrupt Throttling Rate (ints/sec) set to dynamic conservative mode
e1000e: probe of 0000:04:00.0 failed with error -2
We did a lot of research and found few other complaining about
- Problem with wired connection e1000e driver on archlinux.org
- submitted bug with Redhat
- how to configure modprobe to recognize newly installed driver for eth1 on bootup? also didn’t help at all.
- Intel e1000e Ethernet not working
- There were some other peoples reporting that after didn’t fine any solution, left the computer off and worked magically after few days when started again.
Some suggestions:
- Update BIOS (usually is good to have latest version) Not officially available anymore on supermicro.com
- Boot with network cable unplug and connect after system is up
- Use acpi=off on Linux boot loader (if no settings is presented in BIOS)
- Update e1000e Driver from Intel (good to have latest version)
- Try a live CD (just demo version without changing the disk) or try to install Windows and see if the network is recognized.
Working Solution:
Now in our case the server worked fine (we used latest Debian) and was stable. So nothing applied, apart of supposed physical damage of the Network card (two are present and both broken ?!)
We suspected the WOL (Wake on Lan) feature which has been used on these machines, since some time again the SAME type of configuration had the SAME issue. With other words this is second machine having the same issue. On first case we suspected MB and simple replaced. Note also that port led light indication didn’t work (no flash).
Now, disable in BIOS any wake on Lan (or boot from Lan ..) and from your Linux. In our case was /etc/network/interfaces ethernet-wol g
Finally just stop and unplug the server from electricity for several minutes and then restart the machine. This trick worked magically for us.
This solution was also mentioned at Strange ethernet interface issue on new install
Network adapter fails to initialize when system is cold booted
Similar information is confirmed by http://www.aslab.com/support/kb/205.html
Problem:
The e1000e network adapter fails to initialize when the system is cold booted. A message similar to the following can be found in the kernel log:
[ 1.671757] e1000e: Intel(R) PRO/1000 Network Driver - 1.5.1-k
[ 1.671760] e1000e: Copyright(c) 1999 - 2011 Intel Corporation.
[ 1.671803] e1000e 0000:00:19.0: PCI INT A -> GSI 20 (level, low) ->
IRQ 20
[ 1.671815] e1000e 0000:00:19.0: setting latency timer to 64
[ 1.671932] e1000e 0000:00:19.0: irq 50 for MSI/MSI-X
[ 2.721379] e1000e 0000:00:19.0: PCI INT A disabled
[ 2.721385] e1000e: probe of 0000:00:19.0 failed with error -3
If the system is warm booted, the problem does not occur.
Solution: Upgrade the motherboard firmware to the latest version:
Hope it helps some others!