View Issue Details

IDProjectCategoryView StatusLast Update
0014657CentOS-7dracutpublic2018-04-12 15:09
ReporterAlec 
PrioritynormalSeveritymajorReproducibilityalways
Status newResolutionopen 
PlatformOSCent OSOS Version7.3
Product Version7.3.1611 
Target VersionFixed in Version 
Summary0014657: PXE Mount Failed with IPv6 for mount NFS
DescriptionRun PXE mount process on CentOS 7.3 for whole IPv6 load initramdisk and nfs, we found that at the beginning we can load the initial ramdisk successfully, but after the process to mount NFS that we cannot run that over IPv6, we can run only IPv4. Once change the mount parameter to IPv6, we can not run PXE mount successfully, finally it entered the emergency mode.

Loading initial ramdisk ...
EFI stub: Booting Linux Kernel...
EFI stub: Using DTB from configuration table
EFI stub: Exiting boot services and installing virtual address map...
[ 0.000000] Booting Linux on physical CPU 0x0
.
.
.
dracut-initqueue[742]: mount.nfs: an incorrect mount option was specified
[ 59.705231] random: crng init done
[ 142.865597] dracut-initqueue[742]: Warning: dracut-initqueue timeout - starting timeout scripts
[ 143.446997] dracut-initqueue[742]: Warning: dracut-initqueue timeout - starting timeout scripts
.
.
.
Entering emergency mode. Exit the shell to continue.
Type "journalctl" to view system logs.
You might want to save "/run/initramfs/rdsosreport.txt" to a USB stick or /boot
after mounting them and attach it to a bug report.
:/#
Steps To Reproduce1. Run the PXE server with IPv6 setting.
2. Set the grub.cfg as below * root=nfs:[fc00::1]:/opt/nfsroot_centos73
3. Power on the server to boot from PXE mount
4. It boots up normally, but failed at the NFS mount, it showed with message "mount.nfs: an incorrect mount option was specified"
5. Finally boot into emergency mode.

linux (tftp)/netboot/vmlinuz-4.9.0-11.el7.aarch64 pci=pcie_bus_safe pcie_aspm.policy=default cpuidle.off=0 nomodeset root=nfs:[fc00::1]:/opt/nfsroot_centos73:actimeo=1 rw selinux=0 swiotlb=131072 libata.noacpi=1 net.ifnames=0 biosdevname=0
Tags7.3, IPv6
abrt_hash
URL

Activities

kabe

kabe

2018-04-10 09:51

reporter   ~0031590

I couldn't reproduce this on Hyper-V, CentOS 7.4.1708 x86_64.

- try with "rd.debug" kernel option to see dracut processing.

- Is IPv6 address assigned properly?
  try "ip address" on emergency shell.

- Is NFS server reachable?
  try "ping -6 fc00::1" on emergency shell.

- Is NFS server reachable via NFS?
  try "mount -t nfs -o ro [fc00::1]:/opt/nfsroot_centos73 /sysroot" in emergency shell.


Since you seems to be using CentOS 7.3 image,
I recommend updating dracut to the latest ones and
rebuild the initramfs.
Network boot things had improved dramatically on latest dracut package.
Alec

Alec

2018-04-11 11:33

reporter   ~0031598

Try with INITRD from 1708, now without the messages:

dracut-initqueue[742]: mount.nfs: an incorrect mount option was specified

Now still check other things, thanks Kabe.
Alec

Alec

2018-04-11 15:21

reporter   ~0031603

Sorry for previous description as I rechecked that originally something wrong in the dracut setting, so it failed before and not run the mount process, as updated the operation, still with the
message:
dracut-initqueue[742]: mount.nfs: an incorrect mount option was specified

And I also captured the information below, based on the information, it is normal for the network.

dracut:/#
dracut:/# ip address
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP qlen 1000
    link/ether ec:0d:9a:c9:ba:70 brd ff:ff:ff:ff:ff:ff
    inet 192.168.0.161/24 brd 192.168.0.255 scope global dynamic eth0
       valid_lft 258552sec preferred_lft 258552sec
    inet6 fc00::d99/64 scope global dynamic
       valid_lft 258554sec preferred_lft 85754sec
    inet6 fe80::ee0d:9aff:fec9:ba70/64 scope link
       valid_lft forever preferred_lft forever

dracut:/# ping -6 fc00::1
PING fc00::1(fc00::1) 56 data bytes
64 bytes from fc00::1: icmp_seq=1 ttl=64 time=0.238 ms
64 bytes from fc00::1: icmp_seq=2 ttl=64 time=0.179 ms

dracut:/# mount -t nfs -o ro [fc00::1]:/opt/nfsroot_centos73_armrc /sysroot
[ 716.334779] Key type dns_resolver registered
[ 716.366067] NFS: Registering the id_resolver key type
[ 716.376202] Key type id_resolver registered
[ 716.384573] Key type id_legacy registered
dracut:/# cd /sysroot/
dracut:/sysroot# ls
bin dev gcc-build lib media opt root sbin sys usr var.tgz
boot etc home lib64 mnt proc run srv tmp var


But recheck for the rd.debug messages, I found that something strange:

[ 113.069608] dracut-initqueue[1097]: + options=rw,actimeo=1,:,nolock
[ 113.069832] dracut-initqueue[1097]: + mount -t nfs -orw,actimeo=1,:,nolock '[fc00::1]:/opt/nfsroot_centos73_armrc' /sysroot
[ 113.144725] dracut-initqueue[1097]: mount.nfs: an incorrect mount option was specified

Seems that go with one more colon ":" before nolock, but if we follow the same organization for IPv4 is the pass result, may I know if something wrong on that for IPv6. I think my grub setting is ok for without no-needed colon.

IPv6FM-Debug-s.log (1,789,564 bytes)
Alec

Alec

2018-04-12 15:09

reporter   ~0031608

Found that change from nfs to nfs4 and reduce other boot options then we can boot into OS, is there someone encountered that?

Issue History

Date Modified Username Field Change
2018-04-07 05:51 Alec New Issue
2018-04-08 08:36 Alec Tag Attached: 7.3
2018-04-08 08:37 Alec Tag Attached: IPv6
2018-04-10 09:51 kabe Note Added: 0031590
2018-04-11 11:33 Alec Note Added: 0031598
2018-04-11 15:21 Alec File Added: IPv6FM-Debug-s.log
2018-04-11 15:21 Alec Note Added: 0031603
2018-04-12 15:09 Alec Note Added: 0031608