how can a harddisk autocorrects itself all the time while running? (for infine uptim)

frenchn00b · 09-10-2009, 10:00 PM

Helllo

Uptime of 5days, and this ext3 crashes. I know that at reboot itll be funny. testdisk and photorec eventually.
Man, ext3 is fragile, is you dont do tunefs -c and reboot time to time.

So how can I get an huge uptime, not worried, cuz this guy will be able to fix itself while running?

jiml8 · 09-10-2009, 11:13 PM

Ext3 seems pretty robust to me; I never reboot except to install a new kernel, or if a power failure lasts longer than my UPS does.

What do you mean "ext3 crashes"? How do you know this? What are your symptoms?

frenchn00b · 09-20-2009, 07:06 AM

I think we can use smartctl ??

http://www.cyberciti.biz/tips/monito...g-systems.html

I put this into :

Code:

# cat /etc/default/smartmontools
# Defaults for smartmontools initscript (/etc/init.d/smartmontools)
# This is a POSIX shell fragment

# List of devices you want to explicitly enable S.M.A.R.T. for
# Not needed (and not recommended) if the device is monitored by smartd
#enable_smart="/dev/hda /dev/hdb"
enable_smart="/dev/hdf"

# uncomment to start smartd on system startup
start_smartd=yes

# uncomment to pass additional options to smartd on startup
#smartd_opts="--interval=1800"

Code:

# cat /etc/smartd.conf 
/dev/hdf  -m address@gmail.com -M exec /usr/share/smartmontools/smartd-runner

Code:

   Device Boot      Start         End      Blocks   Id  System
/dev/hdf1               1        3039    24410736   83  Linux
/dev/hdf2            3040        3161      979965   82  Linux swap / Solaris
/dev/hdf3            7909       19457    92767342+  83  Linux
/dev/hdf4            3162        3191      240975    b  W95 FAT32

but actually I would like the pc, to check and fix all the time the /dev/hd3 which is the /home

is that possible?

catkin · 09-20-2009, 08:14 AM

Quote:

Originally Posted by frenchn00b

but actually I would like the pc, to check and fix all the time the /dev/hd3 which is the /home

is that possible?

No. At some point physical failures will not be auto-repairable -- unless your HDD is like the VW Beetle in the Woody Allen film. The nearest you can get is hot-swappable storage and some human intervention to replaced failed/failing devices.

frenchn00b · 09-20-2009, 08:47 AM

Quote:

Originally Posted by catkin

No. At some point physical failures will not be auto-repairable -- unless your HDD is like the VW Beetle in the Woody Allen film. The nearest you can get is hot-swappable storage and some human intervention to replaced failed/failing devices.

so how can servers Linux and BSD runs, with huge uptime ?
I dont believe, they can autofix, or use fuse, ... or dont know

well, here freeshell.org
so they daily restart their machine server?

Code:

$ uptime
 1:46PM  up 1 day, 20:36, 43 users, load averages: 0.74, 0.56, 0.57
$

i92guboj · 09-20-2009, 01:40 PM

Smartctl checks the physical integrity of the drive, it doesn't do anything with the fs inside of it. So it really doesn't help with the topic at hand.

Quote:

Originally Posted by frenchn00b

so how can servers Linux and BSD runs, with huge uptime ?
I dont believe, they can autofix, or use fuse, ... or dont know

Raids, redundancy, hot swappable hardware.

frenchn00b · 09-20-2009, 01:50 PM

Quote:

Originally Posted by i92guboj

Smartctl checks the physical integrity of the drive, it doesn't do anything with the fs inside of it. So it really doesn't help with the topic at hand.

Raids, redundancy, hot swappable hardware.

but how does google with their billions of GB, they shall certainly check their drives. How does servers?

i92guboj · 09-20-2009, 01:56 PM

Probably with a combination of raids and clusters, I guess. Google has *lots* of machines, certainly turning one of them off will have zero impact on the services that Google can offer.