Error Message

==== panic kernel thread: 0x2a100875c80 PID: 0 on CPU: 8 ====
cmd: sched(vxdmp:dmp_daemons_loop)
t_procp: 0x19116c0 (proc_sched)
p_as: 0x1914390 (kas)
p_zone: 0x1a1e828 (global)
t_stk: 0x2a100875a90 sp: 0x198ad41 t_stkbase: 0x2a100870000
t_pri: 60 (SYS) pctcpu: 0.000283
t_transience: 10 (TRANSIENT) t_wkld_flags: 0
t_cpupart: 0x198bc50(0) last CPU: 8
idle: 166 ticks (1.66s)
start: Fri Sep 4 11:49:07 2015
age: 1149 seconds (19 minutes 9 seconds)
t_state: TS_ONPROC
t_flag: 0x4808 (T_TALLOCSTK|T_PANIC|T_PUSHPAGE)
t_proc_flag: 0 (none set)
t_schedflag: 3 (TS_LOAD|TS_DONT_SWAP)
t_acflag: 0 (none set)
p_flag: 1 (SSYS)

pc: unix:panicsys+0x48: call unix:setjmp

void unix:panicsys+0x48((const char *)0x10cce50, (va_list)0x2a100875628, (struct regs *)0x198b710, (int)1, 0x4480001607, , , , , , , , 0x10cce50, 0x2a100875628)
unix:vpanic_common+0x78(0x10cce50, 0x2a100875628, 0, 0, 0x2c, 0)
void unix:panic+0x1c((const char *)0x10cce50, (void *)0x1871ef8, 0x704faf28, 0x2a11397bc80, 0x2a100875c80, 0, ...)
unix:mutex_enter(0x704faf28) - frame recycled
int vxio:volinfo_ioctl+0x168((int), (caddr_t)0x2a10087586c, (struct volioctl_info *)0x2a1008757a0, (convent_t)0x1a815d0)
int vxio:volsioctl_real+0x300((voldevno_t), (int), (caddr_t), (int), (cred_t *), (int *))
genunix:cdev_ioctl((dev_t), (int)0x564f4c4a, (intptr_t)0x2a10087586c, (int)0xffffffff80000001, (struct cred *)0x603ac473e38, (int *)0x2a100875868) - frame recycled
void vxdmp:dmp_signal_vold+0x1a8()
void vxdmp:dmp_throttle_paths+0x41c()
void vxdmp:dmp_process_stats+0x278((void *)0)
void vxdmp:dmp_daemons_loop+0x160()
unix:thread_start+4()

In a special scenario of system shutdown or reboot, the DMP (Dynamic MultiPathing) I/O statistic daemon tries to call the ioctl functions in VXIO module which is being unloaded. As a result, the system panics.

A hotfix is now available for this issue in the current version(s) of the product(s) mentioned at the end of this article. Refer to the Hotfix link under Related Documents at the end of this article to obtain the hotfix needed to resolve the issue.

The code is modified to stop the DMP I/O statistic daemon and the DMP restore daemon before the system shutdown or reboot. Additionally, the code is modified to avoid other probes to vxio devices during the shutdown process.

This issue has been fixed in the VxVM 6.0.5.300, 6.1.1.100 and 6.2.1 patches (and above) for Solaris 10 and 11. It has also been included in Infoscale 7.0.1 and 7.1.

Please refer to https://downloads.infoscale.com/ for the latest available patches for these versions.

VxDMP causes panic while rebooting the server.

Applies to:
Solaris 10, 11


 

ETrack: 3677359