Upon transferring some files across server disks I lost connection and the server crashed. Plugging in a Screen showed no response at all. I had to reset the Server, which then boots normally. At first I thought it was Samba, but I verified today that it also happens when issuing move commands via SSH. I’m running Ubuntu Server off a stick on an Asus C60M-I with 2GB non-ECC memory.
Before this all started, SMB transfers would start fast between server drives (30MB/sec, one drive is SATA, the other is mentioned below is eSATA), but would then slow down to an unreasonable speed, at which point I cancelled the transfer.
After reboot it also appears that I’m getting permission denied errors on all my disks. (when non root) Deluge also cannot write files anymore because of this even tough the mount and all contents are owned by nobody with full access.
drwxrwxrwx 5 nobody nogroup 4096 Feb 13 22:40 blastoise drwxrwxrwx 5 nobody nogroup 4096 Feb 12 02:54 charizard drwxrwxrwx 4 nobody nogroup 4096 Feb 12 11:15 magikarp drwxrwxrwx 4 nobody nogroup 4096 Feb 10 02:09 raichu drwxrwxrwx 4 nobody nogroup 4096 Feb 14 03:35 ratata drwxrwxrwx 9 nobody nogroup 4096 Feb 8 19:16 voltorb
Here is the Syslog, which seems to point the finger at the first and only partition (ext4) on sdd:
Then I’ve checked this drive with SmartMonTools which aside from having had a high temperature at one point (still within the max of 60°C), seems to be doing ok:
I’m not sure whether to ask this here or at Server Fault, but any help is appreciated. I’m pretty new at this and I’m figuring things out as I go along.
Usually some share of a volume (5%) is reserved for root. How much filled are the volumes? Maybe more than you think due to the copy / move problem. Have a look at
This reserved share can be set (to 0%, too) by tune2fs:
tune2fs -m 1 /dev/sda1
Force a fsck on all affected volumes in order to rule out file system problems:
e2fsck -fy /dev/sdaX
Have a look at
top when trying to move / copy next time. What happens to the load value?