This issue affects all three servers.
The operating system might panic in function vfs_mountroot if the server is configured to boot from an iSCSI logical unit (LUN) over an Ethernet or InfiniBand network.
If the vfs_mountroot panic occurs, the system console will display output similar to the following:
{0} ok boot net
Boot device: /pci@300/pci@1/network@0 File and args:
1000 Mbps full duplex Link up
1000 Mbps full duplex Link up
1000 Mbps full duplex Link up
SunOS Release 5.11 Version 11.3 64-bit
Copyright (c) 1983, 2016, Oracle and/or its affiliates. All rights reserved.
NOTICE: Configuring iSCSI boot session...
NOTICE: iscsi connection(5) unable to connect to target iqn.1986-03.com.sun:02:118cb23b-920f-49ef-b7a1-b00767a60dfa, target address 192.168.1.1
NOTICE: iscsi connection(5) unable to connect to target iqn.1986-03.com.sun:02:118cb23b-920f-49ef-b7a1-b00767a60dfa, target address 192.168.1.1
WARNING: Failed to configure iSCSI boot session
WARNING: Failed to get iscsi boot path
panic[cpu0]/thread=20012000: vfs_mountroot: cannot mount root
This vfs_mountroot panic will most often be seen during the boot process. However, a similar issue might occur if the system has already booted from an iSCSI logical unit and the iSCSI logical unit becomes temporary unavailable. For example, the iSCSI logical unit could become unavailable as a result of a temporary network congestion or a network failover.
Workarounds
This vfs_mountroot panic issue is usually caused by a slow iSCSI target device or a slow network connection. However, Solaris 11.3 SRU 24 is more vulnerable to this issue due to timing differences in its iSCSI initiator kernel module. If the iSCSI target device or the network is slow, attempt one of the following workarounds to boot the server.
Determine the correct workaround for your server configuration:
If your server contains only one boot environment, go to Apply Workaround to an Installed Boot Environment.
If you freshly installed the Solaris OS on the server, go to Apply Workaround to a Fresh Oracle Solaris OS Installation.
If your SPARC M7 server contains boot environments for Solaris 11.3 prior to SRU 19.5, go to "Select a Boot Environment Earlier than Oracle Solaris 11.3 SRU19.5."
The minimum supported Oracle Solaris OS 11.3 supported on SPARC T8 and SPARC M8 servers is Oracle Solaris 11.3 SRU 24. These servers will not contain boot environments for previous versions of Oracle Solaris 11.3. Follow these steps to apply the workaround to the installed boot environment.
Boot an Oracle Solaris 11.3 installation DVD.
Use the –s option with the boot command. Otherwise you will need to exit the installation program.
{0} ok boot rcdrom -s
Once the server is booted, configure the network interfaces and verify that the server is able to access the iSCSI target IP address.
# ping 192.168.1.1 192.168.1.1 is alive
Configure the iSCSI SendTargets discovery method by providing the discovery address of the iSCSI Target which was used for the iSCSI boot.
# iscsiadm add discovery-address 192.168.1.1 # iscsiadm modify discovery -t enable
Verify that the iSCSI boot logical unit is visible in the format command output.
# format 1. c0t5000CCA02D104088d0 <HGST-H101812SFSUN1.2T-A770-1.09TB> /scsi_vhci/disk@g5000cca02d104088 /dev/chassis/SYS/DBP/HDD0/disk 2. c0t5000CCA02D1032D8d0 <HGST-H101812SFSUN1.2T-A770-1.09TB> solaris /scsi_vhci/disk@g5000cca02d1032d8 /dev/chassis/SYS/DBP/HDD1/disk 3. c0t5000CCA02D102FF8d0 <SUN1.2T cyl 48638 alt 2 hd 255 sec 189> /scsi_vhci/disk@g5000cca02d102ff8 /dev/chassis/SYS/DBP/HDD2/disk 4. c0t5000CCA02D103B20d0 <HGST-H101812SFSUN1.2T-A770-1.09TB> solaris /scsi_vhci/disk@g5000cca02d103b20 /dev/chassis/SYS/DBP/HDD3/disk 5. c0t60000000000000000000000000009999d0 <SUN-COMSTAR-1.0-60.00GB> /scsi_vhci/ssd@g60000000000000000000000000009999 Specify disk (enter its number): 5 selecting c0t60000000000000000000000000009999d0 [disk formatted]
Import the root pool and specify an alternate mount point.
# zpool import -R /a -d /dev/dsk/c0t60000000000000000000000000009999d0s0 rpool
Verify that the iscsi-rpool has been successfully imported.
# zpool list NAME SIZE ALLOC FREE CAP DEDUP HEALTH ALTROOT rpool 59.5G 48.9G 10.6G 82% 1.00x ONLINE /a
Display the boot environments on the imported rpool.
# beadm list be_find_current_be: failed to find current BE name be_find_current_be: failed to find current BE name BE Flags Mountpoint Space Policy Created -- ----- ---------- ----- ------ ------- S11.3-SRU18.6 - - 32.82M static 2027-07-03 13:57 S11.3-SRU22.3 R - 6.42G static 2027-07-03 14:03 solaris - - 99.75M static 2027-07-03 13:40
Create temporary mount point and mount the selected boot environment.
# mkdir /a/mnt # zfs mount -o mountpoint=/a/mnt/ rpool/ROOT/S11.3-SRU22.3
Apply the workaround.
# echo "set iscsi:iscsi_conn_first_login_max=20" >> /a/mnt/etc/system # echo "set iscsi:iscsi_conn_first_interval_max=10" >> /a/mnt/etc/system
Update boot archive in the mounted boot environment to activate the workaround during next reboot.
# bootadm update-archive -v -R /a/mnt
Reboot the system and boot from iSCSI boot device.
# shutdown -y -i0 -g0
{0} ok boot net
Follow these steps to apply the workaround to a freshly installed boot environment.
Repeat the installation, but exit the installation program shortly before the system reboots.
Installation Complete The installation of Oracle Solaris has completed successfully. Reboot to start the newly installed software or Quit if you wish to perform additional tasks before rebooting. The installation log is available at /system/volatile/install_log. After reboot it can be found at /var/log/install/install_log. Esc-4_View Log Esc-7_Halt Esc-8_Reboot Esc-9_Quit
Press the Esc-9 key sequence to exit the installation program and then press 3 to enter the shell.
Welcome to the Oracle Solaris installation menu
1 Install Oracle Solaris
2 Install Additional Drivers
3 Shell
4 Terminal type (currently xterm)
5 Reboot
Please enter a number [1]:3
Display the boot environments.
# beadm list be_find_current_be: failed to find current BE name be_find_current_be: failed to find current BE name BE Flags Mountpoint Space Policy Created -- ----- ---------- ----- ------ ------- solaris R - 2.29G static 2027-07-04 14:46
Mount the newly installed boot environment.
# beadm mount solaris /a
Apply the workaround.
# echo "set iscsi:iscsi_conn_first_login_max=20" >> /a/etc/system # echo "set iscsi:iscsi_conn_first_interval_max=10" >> /a/etc/system
Update boot archive in the mounted boot environment to activate the workaround during next reboot.
# bootadm update-archive -v -R /a
Apply the workaround.
# echo "set iscsi:iscsi_conn_first_login_max=20" >> /etc/system # echo "set iscsi:iscsi_conn_first_interval_max=10" >> /etc/system
Unmount the boot environment and reboot the system.
# beadm umount solaris # reboot
SPARC T8 and SPARC M8-8 servers will never contain boot environments for Oracle Solaris 11.3 versions earlier than SRU 24. If you have a SPARC M7 server that contains earlier boot environments, follow these steps to boot a boot environment earlier than Oracle Solaris 11.3 SRU 19.5 in order to apply the workaround to the Oracle Solaris OS boot environment affected by this vfs_mountroot panic issue.
Select boot environment older than S11.3 SRU 19.5.
{0} ok boot net -L
Boot device: /pci@300/pci@1/network@0 File and args: -L
1000 Mbps full duplex Link up
1000 Mbps full duplex Link up
1000 Mbps full duplex Link up
1 Oracle Solaris 11.3 SPARC
2 S11.3-SRU18.6
3 S11.3-SRU22.3
Select environment to boot: [ 1 - 3 ]: 2
To boot the selected entry, invoke:
boot [<root-device>] -Z rpool/ROOT/S11.3-SRU18.6
Program terminated
Select and boot to a boot environment earlier than Oracle Solaris 11.3 SRU 19.5.
{0} ok boot net -Z rpool/ROOT/S11.3-SRU18.6
List the existing boot environments.
# beadm list BE Flags Mountpoint Space Policy Created -- ----- ---------- ----- ------ ------- S11.3-SRU18.6 N / 33.32M static 2027-07-03 15:57 S11.3-SRU22.3 R - 6.40G static 2027-07-03 16:03 solaris - - 99.75M static 2027-07-03 15:40
Mount the boot environment which exhibits the panic.
# beadm mount S11.3-SRU22.3 /mnt
Apply the workaround.
# echo "set iscsi:iscsi_conn_first_login_max=20" >> /mnt/etc/system # echo "set iscsi:iscsi_conn_first_interval_max=10" >> /mnt/etc/system
Update boot archive in the mounted boot environment to activate the workaround during next reboot.
# bootadm update-archive -v -R /mnt
Unmount the boot environment.
# beadm umount S11.3-SRU22.3
Reboot the server to the new boot environment.