[关闭]
@cdmonkey 2017-05-11T02:03:19.000000Z 字数 8276 阅读 1805

smartmontools

Nagios

https://github.com/v-zhuravlev/zbx-smartctl

一、简介

SMART: Self Monitoring Analysis and Reporting Technology

Download: http://www.smartmontools.org/wiki/Download
Linux.cn: http://linux.cn/article-4461-weixin.html
http://my.oschina.net/u/877567/blog/307336

原创硬盘检测：http://czmmiao.iteye.com/blog/1058215

服务器硬盘是最容易出现问题及发生故障的部件，而且硬盘中存储着大量重要的数据，万一出现故障所造成的损失也是无法估计的，轻则需要化费大量的时间与精力去做数据恢复，重则硬盘报废，里面重要的数据也无法完全恢复，所以对硬盘健康状态的检测，就尤其重要了。为了避免遇到硬盘损坏的情况，用户可使用该软件包程序，它通过使用自身监测、分析及报告三种技术来实施监管存储硬件。

二、安装

安装依赖包：

  1. yum install gcc gcc-c++
  1. [root@oadb tools]# tar zxvf smartmontools-6.4.tar.gz
  2. [root@oadb tools]# cd smartmontools-6.4
  3. [root@oadb smartmontools-6.4]# ./configure --prefix=/usr/local --without-selinux
  4. make && make install
  5. ------------------
  6. # Yum install:
  7. yum install -y smartmontools

三、使用

1. smartctl

  1. # View the version information:
  2. [root@oadb ~]# smartctl -V
  3. smartctl 6.4 2015-06-04 r4109 [x86_64-linux-2.6.18-308.el5] (local build)
  4. Copyright (C) 2002-15, Bruce Allen, Christian Franke, www.smartmontools.org
  5. ...

使用一台配有八块硬盘的服务器进行测试：

  1. [root@kvm-test ~]# smartctl --scan-open
  2. /dev/sda -d scsi # /dev/sda, SCSI device
  3. /dev/bus/6 -d megaraid,0 # /dev/bus/6 [megaraid_disk_00], SCSI device
  4. /dev/bus/6 -d megaraid,1 # /dev/bus/6 [megaraid_disk_01], SCSI device
  5. /dev/bus/6 -d megaraid,2 # /dev/bus/6 [megaraid_disk_02], SCSI device
  6. /dev/bus/6 -d megaraid,3 # /dev/bus/6 [megaraid_disk_03], SCSI device
  7. /dev/bus/6 -d megaraid,4 # /dev/bus/6 [megaraid_disk_04], SCSI device
  8. /dev/bus/6 -d megaraid,5 # /dev/bus/6 [megaraid_disk_05], SCSI device
  9. /dev/bus/6 -d megaraid,6 # /dev/bus/6 [megaraid_disk_06], SCSI device
  10. /dev/bus/6 -d megaraid,7 # /dev/bus/6 [megaraid_disk_07], SCSI device

其中：“megaraid”应该是阵列卡芯片名称，megaraid,0中的0代表的是于“megaraid”中的物理盘编号。

取得设备的“smart”信息：

  1. [root@kvm-test ~]# smartctl --all /dev/sda
  2. smartctl 6.5 2016-05-07 r4318 [x86_64-linux-2.6.32-573.el6.x86_64] (local build)
  3. Copyright (C) 2002-16, Bruce Allen, Christian Franke, www.smartmontools.org
  5. === START OF INFORMATION SECTION ===
  6. Vendor:               DELL
  7. Product:              PERC H710
  8. Revision:             3.13
  9. User Capacity:        2,096,078,258,176 bytes [2.09 TB]
  10. Logical block size:   512 bytes
  11. Logical Unit id:      0x6c81f660e79791001ec4a6da05d41165
  12. Serial number:        006511d405daa6c41e009197e760f681
  13. Device type:          disk
  14. Local Time is:        Wed Apr 19 13:50:19 2017 CST
  15. SMART support is:     Unavailable - device lacks SMART capability.
  17. === START OF READ SMART DATA SECTION ===
  18. Current Drive Temperature:     0 C
  19. Drive Trip Temperature:        0 C
  21. Error Counter logging not supported
  23. Device does not support Self Test logging

能够看到，阵列卡本身并不支持“SMART”功能。再具体看下单块硬盘的情况：

  1. [root@kvm-test ~]# smartctl -i -d megaraid,0 /dev/sda
  2. smartctl 6.5 2016-05-07 r4318 [x86_64-linux-2.6.32-573.el6.x86_64] (local build)
  3. Copyright (C) 2002-16, Bruce Allen, Christian Franke, www.smartmontools.org
  5. === START OF INFORMATION SECTION ===
  6. Vendor:               SEAGATE
  7. Product:              ST9300653SS
  8. Revision:             YS0A
  9. Compliance:           SPC-4
  10. User Capacity:        300,000,000,000 bytes [300 GB]
  11. Logical block size:   512 bytes
  12. Rotation Rate:        15000 rpm
  13. Form Factor:          2.5 inches
  14. Logical Unit id:      0x5000c500716ff47f
  15. Serial number:        6XN5SGF1
  16. Device type:          disk
  17. Transport protocol:   SAS (SPL-3)
  18. Local Time is:        Wed Apr 19 14:49:20 2017 CST
  19. SMART support is:     Available - device has SMART capability.
  20. SMART support is:     Enabled
  21. Temperature Warning:  Disabled or Not Supported

能够看到，硬盘设备支持“SMART”功能。

  1. [root@oadb ~]# smartctl -i /dev/sda
  2. smartctl 6.4 2015-06-04 r4109 [x86_64-linux-2.6.18-308.el5] (local build)
  3. Copyright (C) 2002-15, Bruce Allen, Christian Franke, www.smartmontools.org
  5. Smartctl open device: /dev/sda failed: DELL or MegaRaid controller, please try adding '-d megaraid,N'
  1. # View disk information:
  2. [root@oadb ~]# smartctl -i -d megaraid,0 /dev/sda
  3. smartctl 6.4 2015-06-04 r4109 [x86_64-linux-2.6.18-308.el5] (local build)
  4. Copyright (C) 2002-15, Bruce Allen, Christian Franke, www.smartmontools.org
  6. === START OF INFORMATION SECTION ===
  7. Vendor:               SEAGATE
  8. Product:              ST9300653SS
  9. Revision:             YS09
  10. Compliance:           SPC-4
  11. User Capacity:        300,000,000,000 bytes [300 GB]
  12. Logical block size:   512 bytes
  13. Rotation Rate:        15000 rpm
  14. Form Factor:          2.5 inches
  15. Logical Unit id:      0x5000c5005fb0745b
  16. Serial number:        6XN33ER0
  17. Device type:          disk
  18. Transport protocol:   SAS (SPL-3)
  19. Local Time is:        Thu Oct 22 19:02:55 2015 CST
  20. SMART support is:     Available - device has SMART capability.
  21. SMART support is:     Enabled
  22. Temperature Warning:  Disabled or Not Supported

查看设备的健康状态：

  1. # Check the hard disk health status:
  2. [root@oadb ~]# smartctl -H /dev/sda -d megaraid,0
  3. smartctl 6.4 2015-06-04 r4109 [x86_64-linux-2.6.18-308.el5] (local build)
  4. Copyright (C) 2002-15, Bruce Allen, Christian Franke, www.smartmontools.org
  6. === START OF READ SMART DATA SECTION ===
  7. SMART Health Status: OK

使用SMART只能报告硬盘已经不再健康，可报警后还能继续运行多久无法确定，通常，SMART报警参数是有预留的，硬盘出现报错信息后，并不会当场就损坏，一般能坚持一段时间，有的硬盘报错后还继续跑了好几年，有的硬盘报错后几天就无法使用了，千万不能存侥幸心理。

  1. # Check the hard disk error log:
  2. [root@oadb ~]# smartctl -l error -d megaraid,0 /dev/sda
  3. smartctl 6.4 2015-06-04 r4109 [x86_64-linux-2.6.18-308.el5] (local build)
  4. Copyright (C) 2002-15, Bruce Allen, Christian Franke, www.smartmontools.org
  6. === START OF READ SMART DATA SECTION ===
  7. Error counter log:
  8.            Errors Corrected by           Total   Correction     Gigabytes    Total
  9.                ECC          rereads/    errors   algorithm      processed    uncorrected
  10.            fast | delayed   rewrites  corrected  invocations   [10^9 bytes]  errors
  11. read:   931486797        0         0  931486797          0      13847.275           0
  12. write:         0        0         0         0          0       5962.098           0
  13. verify: 3042089148        0         0  3042089148          0      18667.773           0
  15. Non-medium error count:  2
  1. [root@oadb ~]# smartctl -A -d megaraid,0 /dev/sda
  2. smartctl 6.4 2015-06-04 r4109 [x86_64-linux-2.6.18-308.el5] (local build)
  3. Copyright (C) 2002-15, Bruce Allen, Christian Franke, www.smartmontools.org
  5. === START OF READ SMART DATA SECTION ===
  6. Current Drive Temperature:     31 C
  7. Drive Trip Temperature:        68 C
  9. Manufactured in week 10 of year 2013
  10. Specified cycle count over device lifetime:  10000
  11. Accumulated start-stop cycles:  26
  12. Elements in grown defect list: 0
  14. Vendor (Seagate) cache information
  15.   Blocks sent to initiator = 3753898349
  16.   Blocks received from initiator = 3003303195
  17.   Blocks read from cache and sent to initiator = 2259987377
  18.   Number of read and write commands whose size <= segment size = 40109835
  19.   Number of read and write commands whose size > segment size = 190
  21. Vendor (Seagate/Hitachi) factory information
  22.   number of hours powered up = 16559.07
  23.   number of minutes until next internal SMART test = 19
  1. [root@PBSSES01 ~]# smartctl -a /dev/bus/0 -d megaraid,3
  2. smartctl 6.5 2016-05-07 r4318 [x86_64-linux-2.6.32-573.el6.x86_64] (local build)
  3. Copyright (C) 2002-16, Bruce Allen, Christian Franke, www.smartmontools.org
  5. === START OF INFORMATION SECTION ===
  6. Vendor:               SEAGATE
  7. Product:              ST600MP0005
  8. Revision:             VT31
  9. Compliance:           SPC-4
  10. User Capacity:        600,127,266,816 bytes [600 GB]
  11. Logical block size:   512 bytes
  12. Formatted with type 2 protection
  13. LU is fully provisioned
  14. Rotation Rate:        15000 rpm
  15. Form Factor:          2.5 inches
  16. Logical Unit id:      0x5000c5009599aff7
  17. Serial number:        S7M0WHMN
  18. Device type:          disk
  19. Transport protocol:   SAS (SPL-3)
  20. Local Time is:        Thu Apr 20 14:24:01 2017 CST
  21. SMART support is:     Available - device has SMART capability.
  22. SMART support is:     Enabled
  23. Temperature Warning:  Disabled or Not Supported
  25. === START OF READ SMART DATA SECTION ===
  26. SMART Health Status: OK
  28. Current Drive Temperature:     34 C
  29. Drive Trip Temperature:        60 C
  31. Manufactured in week 48 of year 2015
  32. Specified cycle count over device lifetime:  10000
  33. Accumulated start-stop cycles:  23
  34. Specified load-unload count over device lifetime:  300000
  35. Accumulated load-unload cycles:  489
  36. Elements in grown defect list: 0     # 传说中这里是标识出是否有坏道的地方，俗称成长坏道。
  38. Vendor (Seagate) cache information
  39.   Blocks sent to initiator = 290310405
  40.   Blocks received from initiator = 583431357
  41.   Blocks read from cache and sent to initiator = 1944507
  42.   Number of read and write commands whose size <= segment size = 6490035
  43.   Number of read and write commands whose size > segment size = 51
  45. Vendor (Seagate/Hitachi) factory information
  46.   number of hours powered up = 10990.77
  47.   number of minutes until next internal SMART test = 7
  49. Error counter log:
  50.            Errors Corrected by           Total   Correction     Gigabytes    Total
  51.                ECC          rereads/    errors   algorithm      processed    uncorrected
  52.            fast | delayed   rewrites  corrected  invocations   [10^9 bytes]  errors
  53. read:   2472572503        0         0  2472572503          0        728.491           0
  54. write:         0        0         0         0          0        315.816           0
  55. verify: 597797863        0         0  597797863          0      39623.364           0
  57. Non-medium error count:      2  # 非介质错误。意思是说不是盘的问题，一般是电缆、传输、校验问题，可忽略。
  59. SMART Self-test log
  60. Num  Test              Status                 segment  LifeTime  LBA_first_err [SK ASC ASQ]
  61.      Description                              number   (hours)
  62. # 1  Background short  Completed                  80       3                 - [-   -    -]
  63. # 2  Reserved(7)       Completed                  64       3                 - [-   -    -]
  65. Long (extended) Self Test duration: 3360 seconds [56.0 minutes]

通过修改以上的 smartmontools 的设置文件定期对硬盘进行健康检测，如同给人定期体检一样，体检通过了并不代表就没病（很多疾病用体检的设备都查不到），所以这也符合“Google”的硬盘报告所说的情况，所有损坏的情况中只有60%可以被SMART检测到，所以不能完全依赖其检测结果。

Error log

  1. [root@PBSSES01 ~]# smartctl -l selftest /dev/bus/0 -d megaraid,3  
  2. smartctl 6.5 2016-05-07 r4318 [x86_64-linux-2.6.32-573.el6.x86_64] (local build)
  3. Copyright (C) 2002-16, Bruce Allen, Christian Franke, www.smartmontools.org
  5. === START OF READ SMART DATA SECTION ===
  6. SMART Self-test log
  7. Num  Test              Status                 segment  LifeTime  LBA_first_err [SK ASC ASQ]
  8.      Description                              number   (hours)
  9. # 1  Background short  Completed                  80       3                 - [-   -    -]
  10. # 2  Reserved(7)       Completed                  64       3                 - [-   -    -]
  12. Long (extended) Self Test duration: 3360 seconds [56.0 minutes]

2. smartd

上面的smartctl看起来是个非常不错的工具，可每次都通过手动运行着实麻烦，如果能够以指定的间隔运行，同时又能通知运维人员其检测结果，那不是更好吗？其实该功能已经有了：通过smartd后台服务发就能够实施该功能。

  1. [root@oadb ~]# vim /etc/smartd.conf
  2. # Sample configuration file for smartd. See man smartd.conf.
  3. DEVICESCAN -H -m root
添加新批注
在作者公开此批注前，只有你和作者可见。
回复批注