Bug#627883: initramfs-tools: uses wrong NIC for DHCP, persistent-net invents "eth1-eth0"?
Hi, m...@linux.it (Marco d'Itri) writes: > On May 25, Timo Juhani Lindfors wrote: > >> From this shell I see that the right NIC (i.e. the one with MAC of >> 00:25:90:36:c0:d2) has the name "eth1-eth0" and "eth0" is a wrong NIC. I > Can you still reproduce this with 204-9? I no longer have access to the test setup so I can't test this unfortunately. -Timo -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#627883: initramfs-tools: uses wrong NIC for DHCP, persistent-net invents "eth1-eth0"?
On May 25, Timo Juhani Lindfors wrote: > From this shell I see that the right NIC (i.e. the one with MAC of > 00:25:90:36:c0:d2) has the name "eth1-eth0" and "eth0" is a wrong NIC. I Can you still reproduce this with 204-9? -- ciao, Marco signature.asc Description: Digital signature
Bug#627883: initramfs-tools: uses wrong NIC for DHCP, persistent-net invents "eth1-eth0"?
m...@linux.it (Marco d'Itri) writes: > Please report the output of "ip link". At what time? When I'm in the beginning of do_nfsmount? When I workaround the problem by using "net0" instead of "eth0" I see oper@watt:~$ ip link 1: lo: mtu 16436 qdisc noqueue state UNKNOWN link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 2: eth2: mtu 1500 qdisc noop state DOWN qlen 1000 link/ether 00:1b:21:a1:d7:83 brd ff:ff:ff:ff:ff:ff 3: net0: mtu 1500 qdisc mq state UP qlen 1000 link/ether 00:25:90:36:c0:d2 brd ff:ff:ff:ff:ff:ff 4: eth3: mtu 1500 qdisc noop state DOWN qlen 1000 link/ether 00:1b:21:a1:d7:82 brd ff:ff:ff:ff:ff:ff 5: eth1: mtu 1500 qdisc noop state DOWN qlen 1000 link/ether 00:25:90:36:c0:d3 brd ff:ff:ff:ff:ff:ff when the system is fully booted up. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#627883: initramfs-tools: uses wrong NIC for DHCP, persistent-net invents "eth1-eth0"?
Please report the output of "ip link". -- ciao, Marco signature.asc Description: Digital signature
Bug#627883: initramfs-tools: uses wrong NIC for DHCP, persistent-net invents "eth1-eth0"?
m...@linux.it (Marco d'Itri) writes: > Did you rebuild the initramfs as explained in README.Debian? > Search for NEED_PERSISTENT_NET. Yes it detected the scenario automatically since fstab has /dev/nfs / nfserrors=remount-ro 0 0 which is matched by /usr/share/initramfs-tools/hooks/udev: if egrep -q '^/dev/(nbd|nd[a-z]|etherd/e)[^[:space:]]+[[:space:]]+/[[:space:]]' /etc/fstab; then The right 70-persistent-net.rules is inside the initramfs image: oper@watt:~$ zcat /boot/initrd.img-2.6.32-5-amd64 | cpio -i --to-stdout lib/udev/rules.d/70-persistent-net.rules # lindi: ethernet port #1 (BIOS boots using this) SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:25:90:36:c0:d2", ATTR{dev_id}=="0x0", ATTR{type}=="1", KERNEL=="eth*", NAME="eth0" # PCI device 0x8086:0x10ec (ixgbe) SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:1b:21:a1:d7:83", ATTR{dev_id}=="0x0", ATTR{type}=="1", KERNEL=="eth*", NAME="eth2" # PCI device 0x8086:0x10ec (ixgbe) SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:1b:21:a1:d7:82", ATTR{dev_id}=="0x0", ATTR{type}=="1", KERNEL=="eth*", NAME="eth3" # PCI device 0x8086:0x10c9 (igb) SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:25:90:36:c0:d3", ATTR{dev_id}=="0x0", ATTR{type}=="1", KERNEL=="eth*", NAME="eth1" 53853 blocks -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#627883: initramfs-tools: uses wrong NIC for DHCP, persistent-net invents "eth1-eth0"?
On May 25, Timo Juhani Lindfors wrote: > oper@watt:~$ zcat /boot/initrd.img-2.6.32-5-amd64 | cpio -i --to-stdout > lib/udev/rules.d/70-persistent-net.rules Did you rebuild the initramfs as explained in README.Debian? Search for NEED_PERSISTENT_NET. -- ciao, Marco signature.asc Description: Digital signature
Bug#627883: initramfs-tools: uses wrong NIC for DHCP, persistent-net invents "eth1-eth0"?
Package: initramfs-tools Version: 0.98.8 Severity: normal Description === On a network booting machine with oper@watt:~$ zcat /boot/initrd.img-2.6.32-5-amd64 | cpio -i --to-stdout lib/udev/rules.d/70-persistent-net.rules # lindi: ethernet port #1 (BIOS boots using this) SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:25:90:36:c0:d2", ATTR{dev_id}=="0x0", ATTR{type}=="1", KERNEL=="eth*", NAME="eth0" # PCI device 0x8086:0x10ec (ixgbe) SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:1b:21:a1:d7:83", ATTR{dev_id}=="0x0", ATTR{type}=="1", KERNEL=="eth*", NAME="eth2" # PCI device 0x8086:0x10ec (ixgbe) SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:1b:21:a1:d7:82", ATTR{dev_id}=="0x0", ATTR{type}=="1", KERNEL=="eth*", NAME="eth3" # PCI device 0x8086:0x10c9 (igb) SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:25:90:36:c0:d3", ATTR{dev_id}=="0x0", ATTR{type}=="1", KERNEL=="eth*", NAME="eth1" 53853 blocks and $ grep DEVICE= /etc/initramfs-tools/initramfs.conf DEVICE=eth0 I see that ipconfig gets stuck at bootup trying to send DHCP requests over a NIC whose MAC is not 00:25:90:36:c0:d2. Troubleshooting === I placed a call to "/bin/sh" just before configure_networking call in oper@watt:~$ zcat /boot/initrd.img-2.6.32-5-amd64 | cpio -i --to-stdout scripts/nfs | head -n15 # NFS filesystem mounting -*- shell-script -*- # FIXME This needs error checking retry_nr=0 # parse nfs bootargs and mount nfs do_nfsmount() { configure_networking # get nfs root from dhcp if [ "x${NFSROOT}" = "xauto" ]; then # check if server ip is part of dhcp root-path >From this shell I see that the right NIC (i.e. the one with MAC of 00:25:90:36:c0:d2) has the name "eth1-eth0" and "eth0" is a wrong NIC. I can see in source code of udev that it sometimes generates such names but I don't fully understand the logic behind that. I did rmmod ixgbe rmmod igb modprobe igb modprobe ixgbe and suddenly 00:25:90:36:c0:d2 was correctly named eth0. The machine has 36 SATA disks and it takes a while to detect them. Maybe there's a race condition somewhere? As a workaround I'm now calling my 00:25:90:36:c0:d2 NIC "net0" instead of "eth0". -- Package-specific info: -- initramfs sizes -rw-r--r-- 1 root root 9.1M May 23 16:01 /boot/initrd.img-2.6.32-5-amd64 -- /proc/cmdline root=/dev/nfs initrd=watt/initrd.img nfsroot=10.8.8.8:/tftpboot/watt ip=dhcp sysrq_always_enabled break=top BOOT_IMAGE=watt/vmlinuz -- resume RESUME=UUID=d2e19371-27b1-4826-9d93-dd0f4e1f0ea3 -- /proc/filesystems -- lsmod Module Size Used by loop 11799 0 snd_pcm60503 0 snd_timer 15582 1 snd_pcm snd46446 2 snd_pcm,snd_timer soundcore 4598 1 snd snd_page_alloc 6249 1 snd_pcm psmouse49937 0 joydev 8459 0 pcspkr 1699 0 serio_raw 3752 0 i2c_i8017830 0 evdev 7352 4 ioatdma34876 32 i2c_core 15819 1 i2c_i801 button 4650 0 processor 29935 16 nfs 241114 1 lockd 57603 1 nfs fscache29834 1 nfs nfs_acl 2031 1 nfs auth_rpcgss33476 1 nfs sunrpc161541 9 nfs,lockd,nfs_acl,auth_rpcgss sd_mod 29921 0 crc_t10dif 1276 1 sd_mod usbhid 33292 0 hid63257 1 usbhid uhci_hcd 18521 0 ehci_hcd 32081 0 ixgbe 106622 0 usbcore 122386 4 usbhid,uhci_hcd,ehci_hcd mpt2sas87773 0 scsi_transport_sas 19865 1 mpt2sas mdio3096 1 ixgbe scsi_mod 122149 3 sd_mod,mpt2sas,scsi_transport_sas nls_base6377 1 usbcore igb77959 0 dca 3761 3 ioatdma,ixgbe,igb thermal11674 0 thermal_sys11942 2 processor,thermal -- /etc/initramfs-tools/modules -- /etc/kernel-img.conf # Kernel image management overrides # See kernel-img.conf(5) for details do_symlinks = yes do_bootloader = no do_initrd = yes link_in_boot = no -- /etc/initramfs-tools/initramfs.conf MODULES=most BUSYBOX=y KEYMAP=n COMPRESS=gzip BOOT=nfs DEVICE=eth0 NFSROOT=auto -- /etc/initramfs-tools/update-initramfs.conf update_initramfs=yes backup_initramfs=no -- mkinitramfs hooks /etc/initramfs-tools/hooks/: /usr/share/initramfs-tools/hooks: busybox dmsetup keymap klibc mdadm thermal udev -- System Information: Debian Release: 6.0.1 APT prefers stable-updates APT policy: (500, 'stable-updates'), (500, 'stable') Architecture: amd64 (x86_64) Kernel: Linux 2.6.32-5-amd64 (SMP w/16 CPU cores) Locale: LANG=C, LC_CTYPE=en_GB.UTF-8