My now 6 year old plus Virtualisation Server normally boots up from Bios (20 seconds) and then to load it's Operating System (30 seconds) for a total Boot time of under 1 minute.
Not bad for a /server machine/ which is designed to run 24x7 without a break if need be.
So a boot time of over 15 minutes, and, after pressing C-A-Delete at the login window to Desktop in another 20 minutes this morning had me thinking.
Probably, something is very wrong!
Last week whilst fixing another problem I noticed the SMART disk status indicators on this disk looking very unhealthy, now I realise that things have got dramatically worse.
It's interesting to note that this was not a (Windows) Operating System disk issue, just an unrelocatable error on a disk on which backups were stored.
Yet, it's breakdown had caused the issue. How do I know? Well, removal of this faulty disk and a reboot of the system without it put back the boot time to 50 seconds.
Hmm.
Now, of course all Microsoft Windows systems, certainly since XP have a centralised and well organised Application, Security, Setup and System logging system that you can refer to anytime.
So from this we can instantly see the issue.
So I got to thinking how about other Operating Systems manage.
IBM's AIX has a well established but proprietary facility shown using the errpt command.
This article shows you some features
Apple OSX
Since OSX runs UNIX under the covers we can look in the standard syslog area /var/log/system.log to check for disk errors.
OSX also has the diskutil command that can be used to list and check disks.
Centos 7 Linux
Centos runs the /usr/sbin/rsyslogd program writing to /var/log/messages
Android
Writes to a file called /dev/log/main
Back to fixing my problem
- I Removed possibly failing disk
- Tested with HD Tune Pro, shows bad Logical Blocks for sure
- Place a new disk into my Virtualisation Server (Another old Apple branded disk, this time 500GB! . Will surface test for sure ...)

- Testing full surface of disk. It took over 2 hours but I think this time is well spent.


- Re-setup this disk as a target for backups ! :-) Note that your disk is formatted and any drive letter removed. So caution
(After completion you can always put a letter back onto the disk and then use it for other purposes in addition to the backup)
Commentary
I have a huge library of hard disks but the majority of them are now about 30 years old (and working!). These disks are used sparingly but hard disk error rates have improved so dramatically that it is still quite a shock /to the system/ when any modern hard disk failure occurs at home.
Many of us are now moving to SSD disks for reasons of speed and these disks have a known and limited lifetime! [Having said that my last 3 hard disk failures in the last 12 months were from spinning disks.]
So I really should prepare myself for this to be a more frequent occurrence and be prepared to fully recover from it
Summary
My Windows server has an excellent Centralised Error log which I can view graphically.
SMART disk errors get reported there and I should not need to invest in a special "look at my SMART disk errors tool, please buy me" application
HD Tune Pro is however an excellent [chargable] Benchmarking and checking disk tool which also checks SMART, and I would recommend it
Most other OS, today based on UNIX/ LINUX use the traditional syslog facility or variant thereof to receive disk errors that the user must then interpret.
Hard Disks have a finite lifetime and you should make 100% sure that any data you care about has a backup.