在LCD上显示INIT CBF8

木言

1、        启动
        HP-UX Start-up in progress                                  Status
  ---------------------------------------------                                                                ----------------
Starting HP Distributed  Print Service …………………………………[N/A]
Start clock daemon ……………………………………………………….[OK]
Start diagnostic subsystem ………………………………………………..[OK]
Starting hardware predictive ………………………………………………[busy]


Stored message buffer up to panic
8/4.0.0 sdiskrror has occurred!
8/4.3   Refer to the file /etc/rc.log for more information
…..
System Panic:
  9245XB HP-UX (B.10.20) #1:Sun Jun 9 06:31:19 PDT 1996
panic : (display==0xb800, flags==0x0) Data page fault

PC-Offset Stack Trace (read across, most recent is lst):
   0x001e2244    0x001bbba8  0x001bc260        0x00021668        0x000bd0b4 0x000bcf24
   0x000a3f44…………
   0x000b6464        0x00046a50
End of Stack

NOT sync’ing disks (on the ICS) (457 buffers to flush): 457 457  457   457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457  457
  457buffers not flushed(注：这里的457会改变)
345buffers still dirty
It was not possible for the kernel to find a process that caused this crash.


Dumpsys() called
Attempting to dump 268435456 bytes of memory to disk device 0x1f026000
Disk write: mem addr 0x3c5d00, disk addr 0x132d8000, count 2048.
……….
Disk write : mem……………….
Further “disk write:” messages suppressed.
Successfully dumped 267435456 bytes to dev 0x1f026000,offset 314208 Kbytes
此时LCD显示 INIT CBF8
2、自动重新启动后
能够进入多用户模式。
手动重新启动，有如下错误：
copy processor logs to /var/tombstones …………………………………………………..[busy]

System Panic:
  9245XB HP-UX (B.10.20) #1:Sun Jun 9 06:31:19 PDT 1996
panic : (display==0xb800, flags==0x0) Data page fault

PC-Offset Stack Trace (read across, most recent is lst):
   0x001e2244    0x001bbba8 0x000b5b84 0x00022608                0x000a9c14 0x000a9424
   0x000a1a5c 0x000fc7a8 0x00117918  0x000b6464 0x00046a50
End of stack

0 buffers dirty timeout set to 0 seconds

sync’ing disks(0 buffers to flush): (0 buffers to flush)
0buffers not flushed
0 buffers still dirty
It was not possible for the kernel to find a process that caused this crash.

Dumpsys() called
Attempting to dump 268435456 bytes of memory to disk device 0x1f026000
Disk write: mem addr 0x3c5d00, disk addr 0x132d8000, count 2048.
……….
Disk write : mem……………….
Further “disk write:” messages suppressed.
Successfully dumped 267435456 bytes to dev 0x1f026000,offset 314208 Kbytes
此时LCD显示 INIT CBF8

3、让其自动重启三次均报第二种错误，第四次则到login后出现system panic
NOT sync’ing disks
0 buffers not flushed
1525 buffers still dirty
其余重复。
LCD显示由TEST 2040,TEST2070变为INIT C440
                                   SERIAL 1 CPU0

4、同样启动过程中会用/sbin/fsclean 不能clean lvol4 lvol5 lvol6 lvol7 lvol8 ,然后自动运行fsck做respair，并fixed.
然后又错误如1。

环境：HP D CLASS9000+HPUX

木言

我也是十万火急。就没有人理我吗？

看来我的人际关系太差火了。

kisa_lcx

原帖由 "木言" 发表：
我也是十万火急。就没有人理我吗？

看来我的人际关系太差火了。

你的系统有硬件错误了, 估计是磁盘或者IO bus 部分.
CBF8 PDC_IO initialization completed
CBF9 PDCE_HPMC or PDC_IO found unconfigured IOA or bus converter
CBFA Previous HPMC PIM logged, current HPMC can not be logged
CBFB Branching to the OS HPMC handler
CBFC Branch to OS HPMC failed
CBFE HPMC interrupted a TOC
CBFF Nested HPMC occurred

kisa_lcx

C440 Initializing the stable storage console path
观察 8/4.0.0 通道上的硬盘.
可以把/var/tombstones/ts99 的内容发来看看, 你也可以联系HP 的响应中心.
TEST 2040 CACHE_ERR Testing the parity detection circuitry in the
caches.
FLT 2040 CACHE_ERR_HPMC HPMC due to cache error.
FLT 2070 DCACHE_RAM Failure testing the D-cache RAMs.

木言

你说 对了。有TEST 2040
                    TEST 2070.

谢谢！！

Janson-Chen

你的机器出现了HPMC，你联系HP响应中心，他们会分析这个东东。

木言

非常奇怪！

我重新启动了大概15次，每次都是单用户模式，都做fsck，然后我init 2就可以进入多用户了，而且也不提示错误或者重新启动了。

不知道为什么

wdbj

把/var/tombstones/ts99 下产生的HPMC发送给HP，可以具体分析是什么故障。多半是硬件问题。