I have this problem for quite some time now. every few days, 

1. Webgui is unaccesible.

2. ssh to /mnt/user/ will hang putty.

3. samba shares are unavailable.

 

I still can ssh though.

 

I cant issue reboot command.

 

I already force reboot out of frustration.

 

Ive been thinking, since the problem is with the /mnt/user/ folder, maybe force unmount and then remount would solved it.

 

to force unmount

fusermount -uz /mnt/user/

 

but how do I remount?

 

I would like to try this at the next unraid hang.

 

Thanks.

my server crashed again. i have successfully unmount. how do i remount?

how do i stop and start the array from command line?

webgui still doesnt work.

i try opening every disk in /mnt/. I have disk1-disk14. i am unable to open disk10. seems like disk10 is the source of problem. how do i identify disk10? at current state, i also cant reboot probably due to disk10 stalling the system. how do i unmount disk10?

You can get the diagnostics file by typing   diagnostics  on the command line.  It will be written to the logs  folder on the flash drive. 

 

To shut down the server, type powerdown on the command line.  (I believe shutdown will work also.)  You did get the diagnostics file before you entered this command, didn't you????   

I have already reset my server. I believe the problem is caused by one of my disk. Smart already reported a problem with one of my disk. Investigation of the problem is not my interest right now. I only want to know how can i clean reboot without having to validating parity.

 

Thanks.

Goto the config folder.  Open the disk.cfg file in an editor. (make a backup copy first!)  Find this line
 

startArray="yes"

change the yes to a no

 

That will prevent the autostart on the next boot which avoid the parity check. 

Actually my main problem is, how do i clean reboot? Cancelling parity check at next reboot is not really a problem for me. 

 

I mean, if I issue reboot command, nothing happens. Because disk10 is stalling the system. I dont know which process to kill so that I can clean reboot. I cant reboot from webgui. webgui is dead. i cant stop samba. samba in not responsive.

 

What should I do to be able to clean reboot when my disk10 hang?

 

Thanks.

1 hour ago, publicENEMY said:

how do i clean reboot?

 

powerdown -r

or

reboot

 

There are some instructions for doing a clean shutdown from the command line, but they are somewhat dates and don't take into account some of the newer technologies (like shutting down Dockers and VMs).

 

The powerdown script is supposed to have the smarts to cleanly shut down the array and is a better option.

 

But if you are having hardware issues with a drive in the away, all bets are off in being able to cleanly shut down.

 

9 hours ago, publicENEMY said:

I have already reset my server. I believe the problem is caused by one of my disk. Smart already reported a problem with one of my disk. Investigation of the problem is not my interest right now. I only want to know how can i clean reboot without having to validating parity.

 

Thanks.

 

A noisy child demands your attention / interest. And this child appears to be screaming its head off! 

I have replaced all problematic drives. Currently in the process of migrating from rfs to xfs. 

 

Im curious. When is the time the power load is the maximum? During startup only? Can parity check load psu to the max? What happens when power is not enough? Im just trying to rule out my psu as potential problem.

7 minutes ago, publicENEMY said:

When is the time the power load is the maximum? During startup only?

Usually the peak power draw will be when all of the drives are being spun up simultaneously.  This will happen at several times during routine operation.  Parity checks and Array shutdown are the two principal one that usually come to mind.  Inrush currents for each hard drive start-up are in the range of 2 to 3 amperes depending on the particular drive.  While this peak requirement is only for a few milliseconds, PS's can and have shutdown as a result of the peak rating being exceeded!

 

9 minutes ago, publicENEMY said:

What happens when power is not enough?

Basically, the PS will just shutdown.  Please note that you only need to exceed the rating for a couple of milliseconds for this to occur.  (Semiconductors fail every quickly once you exceed their max ratings because of the small amount of mass involved.  Thus, the protection circuits are designed to do their thing instantaneously.)    Please note that there is also a peak power rating on the PS itself and there are a lot of PS (for example) with say a 400 Watt rating which have have a 33 ampere rating on the 12V rail.  If you do the math, you will see if you have a real world server MB that nominally idles at 25W on this supply and you actually pull 32 Amperes from that buss to spin up a set of hard drives, the PS will shutdown!

I guess, i dont have PSU problem. Currently halfway converting from reiserfs to xfs.