since i cant get my system out of standby/suspention, im looking for a command to disable it all.

i still want the drives to spin up and down, but the cpu must remain on all the time.

 

dunno if killall acpid will do the trick, cos i dont know whats behind it all (Tom?)

 

 

the thing i dont want, is that i need to send a WOL packed (which doenst work here), every time my system is idle for more then 1 hour (drive spindown time)

i want it to be available for the network when it is needed, without any external wakeup commands.

 

and i cant get it out of standby/suspention because everything is powered down exept the LAN (even ps/2 keyboard and usb).

seems liky S3 (or can it be S1), but i set my bios to only S1

The acpi subsystem just provides hooks to let linux issue a 'soft power off', ie, the same thing that happens when you push a front-panel power switch; it's not currently used for anything else.

 

If you click 'Power down' on the Management Utility, then you should be able to power back up a few ways:

 

a) Push the front panel power switch.  If no front panel try using a screw-driver to short the two pins on the motherboard where the front panel power switch connects.  This will test if your m/b is indeed in S5 suspend.

 

b) Turn the power supply switch off then back on again.  If you have dual supplies as in the MD-1200, only need to turn off/on the top supply & bottom supply will come on when top one does.

 

c) Send the WOL packet to your network adaptors MAC address.

 

Some bios have built-in power management - I suggest you turn that all off.

 

Other than that, if you never click 'Power down', then system will never try to do any other kind of bios-supported suspend.

 

So drives will spin-down after inactivity, and linux executes a 'stop' instruction when it has nothing to do, effectively putting it in S1 suspend - next interrupt will cause CPU to execute instructions again.  You don't have to do anything to enable this behavior - it's built in and has nothing to do with acpi.

ok i get it, thats what i want 

 

but, when all my drives spin down, the system is unreachable

 

no management //tower,

no network \\tower,

no monitor,

no usb,

no ps\2 keyboard/mouse

fans off, no heat produced

 

only ethernet is still blinking, but WOL packet doesnt work (tried numerous programs, MAC/ip adres triple checked)

 

my motherboard (asus i915Ga-HFS) gives me the options S1, S3 or S1 & S3 (no disable), so i set it to S1.

 

im totaly lost on this, it should work, but it doesnt

 

just a question: "next interrupt will cause CPU to execute instructions again" -> i dont have to manually wake it up?, when i network \\tower it will show up without any problems?

When you say, "when all my drives spin down, the system is unreachable" - what do you mean by that?  Is this the result of clicking 'Power down' on the Main page of the Management Utility?  Or does your system 'shut down' by itself?  or something else?

 

Clicking on "Manual" on the Aopen site for your motherboard produces a 'Page Not Found' at the moment, so I can't see your bios options.

 

We set up the Intel boards for "S3", but it doesn't make any difference at this point because we don't tell the system to suspend.

well, i set the spindown timer to 1 hour.

when that 1 hour passes, the drives spindown

at the same time the drives are spinning down, i see the power being removed from the systems described above

and seconds there after the system is totally unreachable

 

i am not giving any commands of any sort, its doing it on its own.

 

i have to take the power off the system completly to restart it.

the power button and reset button do nothing at all, at that time (but they both work when the system is booting/running)

what does work is holding the powerbutton for more then 5 secs, then the system is off (not that i can see any visual changes, but then i can start the system again)

 

it surly doesnt shutdown nicely, cos when it's restarted, it does a parity check, or when i do it to much a parity sync.

Hmm, never seen this happen before.  I wonder if you could try this:

 

Go to the settings page and set the spin down delay to 2 hours.  Then open a telnet window and set it up to monitor the syslog using this command:

 

  tail -f /var/log/syslog

 

(Instructions for doing this are here).

 

Then wait and see what happens at 1 hour mark and 2 hour mark.

 

What I'm trying to do here is see if there is a definite correlation between spinning the drives down and the system being shutdown.  The open telnet window will let you capture the syslog, which gets lost when it powers off.

uploaded the manual, since the aopen site isnt the easiest to reach

http://members.chello.nl/whoek10/i915gahfs-ol-e.pdf

 

ok, i did what u asked and here is the result:

 

• booted
  
• stopped parity
  
• set spindown to 2 hours
  
• cleared statistics
  

 

 

 

Tower login: root

Linux 2.4.33.

root@Tower:~# tail -f /var/log/syslog

Oct 05 10:07:24 Tower emhttp[1052]: user share: Movies

Oct 05 10:07:24 Tower emhttp[1052]: user share: Series

Oct 05 10:07:24 Tower emhttp[1052]: shcmd (23): /usr/sbin/nmbd -D

Oct 05 10:07:24 Tower emhttp[1052]: shcmd (24): /usr/sbin/smbd -D

Oct 05 10:07:47 Tower emhttp[916]: driver cmd: nocheck

Oct 05 10:07:47 Tower kernel: mdcmd (20): nocheck

Oct 05 10:07:47 Tower kernel: md: md_do_sync() got signal ... exiting

Oct 05 10:07:47 Tower kernel: md: sync done. time=76sec rate=11125K/sec

Oct 05 10:07:47 Tower kernel: md: writing superblock to /boot/config/super.dat

Oct 05 10:07:47 Tower kernel: md: recovery thread sync completion status: -4

Oct 05 10:09:20 Tower emhttp[916]: driver cmd: clear

Oct 05 10:09:20 Tower kernel: mdcmd (26): clear

Oct 05 12:09:39 Tower emhttp[972]: shcmd (25): /usr/sbin/hdparm -y /dev/sdb >/de

v/null

Oct 05 12:09:39 Tower emhttp[972]: shcmd (26): /usr/sbin/hdparm -y /dev/sde >/de

v/null

Oct 05 12:09:39 Tower emhttp[972]: shcmd (27): /usr/sbin/hdparm -y /dev/sdc >/de

v/null

Oct 05 12:09:40 Tower emhttp[972]: shcmd (28): /usr/sbin/hdparm -y /dev/sdd >/de

v/null

 

 

after the 2 hours, the drives spindown (like the concole says) and directly there after the system is unreachable  :'(

Nothing unusual about the log.  The entries toward the end which look like this:

 

Oct 05 12:09:39 Tower emhttp[972]: shcmd (25): /usr/sbin/hdparm -y /dev/sdb >/dev/null

 

are logging commands to spin the drives down.  This simply translates to an IDE command to the drive.

 

Thanks for the manual link.  Looking at the manual, there are some funny things having to do with power UP, such as being able to set a timer, etc.  Make sure all this stuff is "OFF".

 

Here is something you can try.  There is a file in the root of the flash called 'syslinux.cfg'.  Edit this file and change this line:

 

append initrd=bzroot ramdisk_size=120000 root=/dev/ram0

 

change to:

 

append initrd=bzroot ramdisk_size=120000 root=/dev/ram0 acpi=off

 

Then reboot the system and see if it still powers off after spinning the drives down.

--

Note: I should add that turning 'acpi' off like this will cause the 'Power down' button on the Main page to not work.

well tried it numerous times, still no luck  :'( :'( :'(

 

im just curious, could this be to a bug in the motherboard bios (there is a apple macbook bug like this)

or is it simthing new in unraid?

 

any options left, short of changing the board?

Disabling ACPI by adding 'acpi=off' on the 'append' line in sysling.cfg definitely disables linux from using any kind of power management functionality.  Also, we've never seen this kind of behavior or had reports from anyone else who has....

 

The only other thing I could think of is to make sure your motherboard has the latest bios revision.

 

As an experiment you coud do this.

1. Go to Settings and set Spin down delay to 'Never'.

2. Open telnet session and spin your drives down manually, one by one.  Use this command:

 

  hdparm -y <identifier>

 

use the identifiers below in your syslog, i.e.,

 

  hdparm -y /dev/sdb

  hdparm -y /dev/sdc

  hdparm -y /dev/sdd

  hdparm -y /dev/sde

 

See if your system powers off immediately after the last command.

 

 

havent thought of that yet

 

just tested it, and as soon as i give the command to the last drive to spindown (doesnt matter which drive is last up)

the system imidiately goes into suspent (or something like it) and is unreachable.

 

when i remove all drive sata cables from the motherboard and then boot, the system runs perfectly (w/o any drives, but no freezing)

 

could it be that the hdparm command somehow generates an error?

maybe a wrong setting for my system

 

havent thought of that yet

 

just tested it, and as soon as i give the command to the last drive to spindown (doesnt matter which drive is last up)

the system imidiately goes into suspent (or something like it) and is unreachable.

 

when i remove all drive sata cables from the motherboard and then boot, the system runs perfectly (w/o any drives, but no freezing)

 

could it be that the hdparm command somehow generates an error?

maybe a wrong setting for my system

 

Or, perhaps the power supply needs a certain minimum load and shuts down when all the drives are spun down?

Do you have a spare power supply handy? Worth a try?

well, ever since im suspecting the hdparm command (for 1 hour now  ), i have been wonder if there is something wrong with the drives

so i did a little test.

 

i have been disconnecting the drives from the motherboard and testing them 1 by 1 with the hdparm -y /dev/sdb command

well, guess what they all work fine, no freezing.

so the drives and sata cables are fine

 

then i did the test with 2 drives connected and shutting them down 1 by 1. same result, all works fine (yey  )

 

next, 3 drives, also fine (omg its a miracle)

 

but when i test it with all 4 drives, the system freezes when i shutdown the last drive.

so it has someting to do with the 4th drive

 

 

i think ill be getting a 4 port controller card tomorrow and test the system with 3 drives on the motherboard and 1 on the controller