System fail

[tombost]

Hello,

 

I'm student and i work into a business, we have one server to backup other servers but this server have some problem with unraid. It's start Friday (24/07) and today same problem, the system is down and when we put a screen into the server we can see that :

We can't access webGUI and samba.

We have 4 unraid servers but just one has this problem, what do you think is that ?

 

Thanks you and sorry for my bad english.

Version 6.0.0

[trurl]

Attach a keyboard. Reboot and select memtest from the boot menu. Let it run overnight.

[RobJ]

It's difficult to say from just that one screen.  It could be a hardware issue, such as memory.  Please see Need help? Read me first!, and try to provide a Diagnostics zip file, if at all possible.  We also need to know more about when that screen appears - is it right after boot, after array start, after some operation, etc?

 

I do recommend doing several passes of Memtest, found on the boot menu, to make sure it's not a problem with faulty RAM.  Edit: trurl beat me!

[tombost]

Hi, ok tomorrow i launch a memtest.

I can't give you syslog because the pc is just freeze.. and no network .. and after reboot syslog is clean.

This screen appear when the server is freeze, i can't do everything just press reset button.

 

[bonienl]

Hi, ok tomorrow i launch a memtest.

I can't give you syslog because the pc is just freeze.. and no network .. and after reboot syslog is clean.

This screen appear when the server is freeze, i can't do everything just press reset button.

 

Connect a monitor + keyboard to the system. It allows you to run commands directly on the system. After a crash run:

 

diagnostics

 

The resulting zip file is stored under /boot/logs. You can copy the zip file from flash and post here.

 

[tombost]

Hello,

 

Memtest OK : 6 pass and i stop it

I attach the diag file. Thanks

 

EDIT : other freeze this night, i put a screen this morning : http://ovh.to/jkT4ho

s-sem42562-diagnostics-20150728-1317.zip

[tombost]

New freeze today ... only when we backup data into the server, screen : http://hpics.li/8fc6d6c

[bonienl]

Hello,

 

Memtest OK : 6 pass and i stop it

I attach the diag file. Thanks

 

EDIT : other freeze this night, i put a screen this morning : http://ovh.to/jkT4ho

 

Several incompatible programs are loaded from the go file. I recommend you copy a virgin go file from the unRAID installation zip file and reboot your system.

[tombost]

I'm really sorry but what is a go file ?

You say i need to download a clean unraid zip and change the file config/go ? Thank you

[bonienl]

I'm really sorry but what is a go file ?

You say i need to download a clean unraid zip and change the file config/go ? Thank you

 

Correct change the file config/go or edit the existing file and delete all but the line with emhttp in it

[Squid]

I'm really sorry but what is a go file ?

You say i need to download a clean unraid zip and change the file config/go ? Thank you

The "go" file is on your flash drive in the config folder

[tombost]

I remove the mails config.

I restart and wait, other things : when I enter SMB with windows it's very slow and log show that :

 

Jul 29 08:44:19 S-SEM42562 kernel: REISERFS warning (device md1): jdm-20002 reiserfs_xattr_get: Invalid hash for xattr (user.SAMBA_PAI) associated with [2 19 0x0 SD]
Jul 29 08:44:19 S-SEM42562 kernel: REISERFS warning (device md1): jdm-20002 reiserfs_xattr_get: Invalid hash for xattr (user.SAMBA_PAI) associated with [2 19 0x0 SD]
Jul 29 08:44:19 S-SEM42562 kernel: REISERFS warning (device md1): jdm-20002 reiserfs_xattr_get: Invalid hash for xattr (user.SAMBA_PAI) associated with [2 19 0x0 SD]
Jul 29 08:44:19 S-SEM42562 kernel: REISERFS warning (device md1): jdm-20002 reiserfs_xattr_get: Invalid hash for xattr (user.SAMBA_PAI) associated with [2 19 0x0 SD]
Jul 29 08:44:19 S-SEM42562 kernel: REISERFS warning (device md1): jdm-20002 reiserfs_xattr_get: Invalid hash for xattr (user.SAMBA_PAI) associated with [2 19 0x0 SD]
Jul 29 08:44:19 S-SEM42562 kernel: REISERFS warning (device md1): jdm-20002 reiserfs_xattr_get: Invalid hash for xattr (user.SAMBA_PAI) associated with [2 19 0x0 SD]
Jul 29 08:44:19 S-SEM42562 kernel: REISERFS warning (device md1): jdm-20002 reiserfs_xattr_get: Invalid hash for xattr (user.SAMBA_PAI) associated with [2 19 0x0 SD]
Jul 29 08:44:19 S-SEM42562 kernel: REISERFS warning (device md1): jdm-20002 reiserfs_xattr_get: Invalid hash for xattr (user.SAMBA_PAI) associated with [2 19 0x0 SD]

 

EDIT : problem not solved : transfer 50GB wait 5 min and it freeze again ..(kernel panic)

SCREEN :

http://hpics.li/c709375

[tombost]

I launch a reiserfsck --check but nothing wrong :

[RobJ]

I'm really sorry but something is still very wrong with that system.  You've tested and passed the memory, normally the prime suspect, so that leaves the motherboard or an addon card or something in the system is incompatible or buggy.  The motherboard BIOS appears to be fairly recent (BIOS V2.8 07/17/2014), but perhaps there's a newer one?  I think we're about out of ideas.  What's bad is that not only is the issue crashing your system, but it's corrupting the Reiser file system, raising the chances that it will corrupt data at some point.

 

I was rather surprised to see such a successful result on the reiserfsck of Disk 1, as it clearly looked bad, in both the recent pic and the older syslog.

 

There's only one small issue that I can see and help with.  Linux is case-sensitive, annoyingly and unjustifiably so (in my opinion).  Because of that, you have 4 shares set up, Backups, backup, Installs, and installs.  Obviously you only wanted 2 of them, but you probably have the folders named one way, and somewhere in one or more configuration settings, you have specified the names in the wrong capitalization.  They do have to be exactly the same.  Blame it on Linux!  You will want to fix the settings, then remove the unwanted shares.

[tombost]

Problem solved :

In the go file we have a eth1 config for backup, but Monday we unplug this connection .... I just disable this config and it work fine (move 2,05TB without crash)

 

Thank you for your help.