我们有一台（相当旧的）基于 HP ML115 G5 AMD 的服务器，在按下电源按钮后 10-15 秒后（我想是在风扇测试期间）自动关闭，在 BIOS POST 发出单声哔声之前。
我们需要一些远程（200 公里）硬件故障诊断方面的帮助。我们的硬件规格如下：
root@linux:~/# dmidecode -t1
# dmidecode 2.12
SMBIOS 2.5 present.
System Information
Manufacturer: HP
Product Name: ProLiant ML115 G5
Serial Number: CZC94743QJ
SKU Number: 470064-894`
root@linux:~/# head -n 30 dmidecode.txt
# dmidecode 2.12
Handle 0x0000, DMI type 0, 24 bytes
BIOS Information
Vendor: HP
Version: O18
Release Date: 07/06/2009
此时，它可以稳定运行。我已设法通过以下方式将其打开：
- 关闭服务器，
- 拔掉电源线五分钟，
- 将其较长的一端放置在地面上，CPU 散热器朝向天花板的方向。
如果我们将其置于标准位置，它就不会像我在开头写的那样打开。完全可重现。
电压/温度/风扇统计数据对我来说看起来不错：
root@linux:~/# ipmitool sdr
POST Error | Not Readable | ns
Memory ECC | Not Readable | ns
ACPI State | 0x01 | ok
PCI Reset | 0x00 | ok
CPU Fan | 1048.88 RPM | ok
Rear Fan | 2107.04 RPM | ok
CPU Diode | 26.50 degrees C | ok
Front Ambient | 19 degrees C | ok
System 12V | 11.93 Volts | ok
System 5V | 5.12 Volts | ok
System AUX 5V | 4.98 Volts | ok
System 3.3V | 3.39 Volts | ok
System AUX 3.3V | 3.33 Volts | ok
CPU Vcore | 1.07 Volts | ok
CPU 12V | 11.82 Volts | ok
HT 1.2V | 1.20 Volts | ok
Mem Vcore | 1.81 Volts | ok
MEM VTT | 0.90 Volts | ok
MCP55 1.5V | 1.50 Volts | ok
MCP55 1.4V | 1.40 Volts | ok
Therm-Trip | 0x00 | ok
CPU Prochot | 0x00 | ok
System Reset | 0x00 | ok
NMI | 0x00 | ok
PCI Error | Not Readable | ns
CPU Socket | 0x01 | ok
LO100 Present | 0x00 | ok
Watchdog | Not Readable | ns
IPMI 事件：
18 | 03/18/2015 | 09:29:46 | Temperature #0x20 | Upper Non-critical going high | Asserted
30 | 03/18/2015 | 09:30:08 | Temperature #0x20 | Upper Critical going high | Asserted
48 | 03/18/2015 | 10:38:59 | Temperature #0x20 | Upper Non-critical going high | Asserted
60 | 03/18/2015 | 10:39:20 | Temperature #0x20 | Upper Critical going high | Asserted
78 | 03/18/2015 | 10:45:26 | Temperature #0x20 | Upper Non-critical going high | Asserted
90 | 03/18/2015 | 10:45:30 | Temperature #0x20 | Upper Non-critical going high | Deasserted
a8 | 03/18/2015 | 10:45:56 | Temperature #0x20 | Upper Non-critical going high | Asserted
c0 | 03/18/2015 | 10:46:12 | Temperature #0x20 | Upper Critical going high | Asserted
d8 | 03/18/2015 | 10:48:42 | Temperature #0x20 | Upper Non-critical going high | Asserted
f0 | 03/18/2015 | 10:48:46 | Temperature #0x20 | Upper Non-critical going high | Deasserted
108 | 03/18/2015 | 10:49:04 | Temperature #0x20 | Upper Non-critical going high | Asserted
120 | 03/18/2015 | 10:49:18 | Temperature #0x20 | Upper Critical going high | Asserted
138 | 03/18/2015 | 10:50:24 | Temperature #0x20 | Upper Non-critical going high | Asserted
150 | 03/18/2015 | 10:50:25 | Temperature #0x20 | Upper Critical going high | Asserted
168 | 03/18/2015 | 10:57:53 | Temperature #0x20 | Upper Non-critical going high | Asserted
180 | 03/18/2015 | 10:57:57 | Temperature #0x20 | Upper Non-critical going high | Deasserted
198 | 03/18/2015 | 10:58:24 | Temperature #0x20 | Upper Non-critical going high | Asserted
1b0 | 03/18/2015 | 10:58:41 | Temperature #0x20 | Upper Critical going high | Asserted
1c8 | 03/18/2015 | 11:14:23 | Temperature #0x20 | Upper Non-critical going high | Asserted
1e0 | 03/18/2015 | 11:15:06 | Temperature #0x20 | Upper Non-critical going high | Deasserted
1f8 | 03/18/2015 | 11:16:33 | Temperature #0x20 | Upper Non-critical going high | Asserted
210 | 03/18/2015 | 11:16:33 | Temperature #0x20 | Upper Critical going high | Asserted
228 | 03/18/2015 | 11:49:12 | Temperature #0x20 | Upper Non-critical going high | Asserted
240 | 03/18/2015 | 11:49:18 | Temperature #0x20 | Upper Non-critical going high | Deasserted
258 | 03/18/2015 | 11:55:45 | Temperature #0x20 | Upper Non-critical going high | Asserted
270 | 03/18/2015 | 11:55:46 | Temperature #0x20 | Upper Non-critical going high | Deasserted
288 | 03/18/2015 | 11:56:32 | Temperature #0x20 | Upper Non-critical going high | Asserted
2a0 | 03/18/2015 | 11:57:06 | Temperature #0x20 | Upper Critical going high | Asserted
2b8 | 03/18/2015 | 12:00:11 | Temperature #0x20 | Upper Non-critical going high | Asserted
2d0 | 03/18/2015 | 12:00:14 | Temperature #0x20 | Upper Non-critical going high | Deasserted
2e8 | 03/18/2015 | 12:00:59 | Temperature #0x20 | Upper Non-critical going high | Asserted
300 | 03/18/2015 | 12:01:34 | Temperature #0x20 | Upper Critical going high | Asserted
318 | 07/06/2009 | 00:00:22 | Fan #0x42 | Upper Critical going high | Asserted
330 | 11/13/2016 | 13:25:47 | Fan #0x41 | Upper Critical going high | Asserted
348 | 11/13/2016 | 13:33:00 | Fan #0x41 | Upper Critical going high | Asserted
360 | 11/13/2016 | 13:33:47 | Fan #0x41 | Upper Critical going high | Asserted
378 | 11/13/2016 | 13:44:58 | Fan #0x41 | Upper Critical going high | Asserted
390 | 11/13/2016 | 13:45:48 | Fan #0x41 | Upper Critical going high | Asserted
3a8 | 11/13/2016 | 13:47:45 | Fan #0x41 | Upper Critical going high | Asserted
3c0 | 12/01/2016 | 17:00:29 | Fan #0x41 | Upper Critical going high | Asserted
3d8 | 12/01/2016 | 17:01:53 | Fan #0x41 | Upper Critical going high | Asserted
3f0 | 12/01/2016 | 17:04:02 | Fan #0x41 | Upper Critical going high | Asserted
408 | 12/01/2016 | 17:31:34 | Fan #0x41 | Upper Critical going high | Asserted
420 | 12/01/2016 | 17:43:42 | Fan #0x41 | Upper Critical going high | Asserted
2016 年 11 月 13 日我第一次遇到这种情况，我认为可能是硬件看门狗，所以我们在 BIOS 中将其禁用。
该服务器有 2x1TB 磁盘，2x3TB（不带光驱）。365 瓦非热插拔、非冗余电源。
现在，我们建议更换盒子，但就我而言，我无法解释为什么会发生这种情况（我认为这是某种机械主板故障）。我想知道您是否还有其他想法。
** 更新，Chopper3 先生问我 是什么意思but CPU one is not standard。因此，原来的 hatsink 已损坏如下：
时间和糟糕的材料选择，塑料本来就不耐压。从那次安装之后，我再也没有在其他盒子里看到过塑料支架……
服务器一直保持在良好的环境中，从未过热，没有受到阳光的直接影响，工作期间无人触碰它。
那是大约 1.5 年前的事了。我们在市场上再也找不到原装 HP 部件了。我们用大 3 倍的部件替换了它，因为当时 AM2 插座并不那么流行。我现在记不清它是否有 2 根信号线加上 VCC 和 GND (4)，就像上面发布的库存一样。它可能只有三根。VCC + GND 和旋转信号 (3)。从那时起，我们多次停电，这种情况从未发生过。
答案1
我认为主板存在故障。例如焊点损坏或边缘部件损坏。我遇到过类似的故障，只需推一下主板，服务器就可以启动，但只要我松开压力，服务器就会因风扇故障而关闭或因 ECC 错误而挂起。
答案2
您的风扇可能出现故障，并且服务器配置为在严重风扇故障时停止。