Can't access drive - Need raid5/linux help

[Jocko]

To do the total size do

Code: Select all

du -ch /md4/shares/1

For restoring lacie firmware, there are 4 other raid devices to check, and we also try to overwrite their raid array by recreating them with a missing disk (sdd) so that at booting Lacie succeeds to assemble them.

to unmount the usb device, just do

Code: Select all

umount /md4

(of course the path of cwd must not be /md4 or a subfolder)

[imt17]

Not totally understanding about the firmware part but will discuss more when the backup part is done.

But, I think there may be an issue with that. I had been checking in a second telnet window the progress/directory creation on the USB. It doesn't seem to have changed in a while. The listing still seems pretty much the same as it did over 40min ago. I then ran the command you gave me to check total size. Well I guess that runs through all off the files in the directory showing size and then the telnet window went all black with a blinking cursor at the bottom left. I scrolled back up to see the last entries where part of the subfolders show weird characters in the name. Wonder if it hit bad spot and froze?


132.0K /md4/shares/2/The Firesign Theatre/Shoes For Industry! The Best Of The Firesign Theatre/.AppleDouble
748.4M /md4/shares/2/The Firesign Theatre/Shoes For Industry! The Best Of The Firesign Theatre
60.0K /md4/shares/2/The Firesign Theatre/How Can You Be In Two Places At Once When You're Not Anywhere At All/.AppleDouble
569.3M /md4/shares/2/The Firesign Theatre/How Can You Be In Two Places At Once When You're Not Anywhere At All
1.3G /md4/shares/2/The Firesign Theatre
4.0K /md4/shares/2/上妻宏光/.AppleDouble
48.0K /md4/shares/2/上妻宏光/AGATSUMA/.AppleDouble
269.7M /md4/shares/2/上妻宏光/AGATSUMA
269.7M /md4/shares/2/上妻宏光
4.0K /md4/shares/2/福原　道å­

[Jocko]

Well as the kernel runs only on the RAM, maybe your last command overloaded the free size.

So close your 2d telnet window. Note: it is better to let quiet the NAS when it runs the backup.

[imt17]

Closed it.

Just hoping that it didn't freeze. Guess I won't know until I go home to check. If I do go home and hear no activity on either the USB or NAS, then what should I do?

[Jocko]

You can try to open temporarily a new telnet window and see if ccp command is already running

Code: Select all

ps

if yes see there are some progress on /usb side

[imt17]

Ran ps

Saw that cup time shows 60:12

I then waited a few min and ran PS again and still 60:12

I then ran
ls -a /usb/5big2_backup/2

This time no response. Like it froze

USB light blinking but not sure anything is happening

[imt17]

Here also is result of PS I just tried again in another window:

root@fvdw-sta-kirkwood:/ # ps
PID USER TIME COMMAND
1 root 0:01 /bin/sh
2 root 0:00 [kthreadd]
3 root 3:53 [ksoftirqd/0]
5 root 0:00 [kworker/0:0H]
7 root 0:13 [rcu_preempt]
8 root 0:00 [rcu_sched]
9 root 0:00 [rcu_bh]
10 root 0:00 [khelper]
11 root 0:00 [kworker/u2:1]
210 root 0:00 [writeback]
212 root 0:00 [bioset]
214 root 0:00 [kblockd]
219 root 0:00 [ata_sff]
230 root 0:00 [khubd]
236 root 0:00 [md]
246 root 0:00 [cfg80211]
262 root 0:00 [rpciod]
271 root 10:00 [kswapd0]
272 root 0:00 [fsnotify_mark]
273 root 0:00 [nfsiod]
277 root 0:00 [xfsalloc]
278 root 0:00 [xfs_mru_cache]
279 root 0:00 [xfslogd]
280 root 0:00 [crypto]
405 root 0:00 [scsi_eh_0]
406 root 0:00 [scsi_tmf_0]
409 root 0:00 [scsi_eh_1]
410 root 0:00 [scsi_tmf_1]
414 root 4:58 [kworker/u2:3]
420 root 0:00 [smflush]
428 root 0:00 [spi0]
476 root 0:00 [aoe_tx0]
477 root 0:00 [aoe_ktio0]
500 root 0:00 [raid5wq]
502 root 0:00 [dm_bufio_cache]
503 root 0:00 [kmpathd]
504 root 0:00 [kmpath_handlerd]
522 root 0:00 [mv_crypto]
532 root 0:00 [ipv6_addrconf]
538 root 0:00 [deferwq]
694 root 0:00 /sbin/discoverd
705 root 0:02 /sbin/telnetd
709 root 0:00 /sbin/buttons-nwsp2
1236 root 0:00 [devnod]
1747 root 0:00 -sh
1803 root 0:36 [kworker/0:1H]
2119 root 0:00 [bioset]
2120 root 20:12 [md4_raid5]
2134 root 0:00 [xfs-data/md4]
2135 root 0:00 [xfs-conv/md4]
2136 root 0:00 [xfs-cil/md4]
2137 root 0:02 [xfsaild/md4]
2932 root 0:00 [scsi_eh_3]
2933 root 0:00 [scsi_tmf_3]
2934 root 6:02 [usb-storage]
3142 root 60:12 /bin/ccp -r /md4/shares/2 /usb/5big2_backup
3143 root 0:00 [kworker/u2:0]
3434 root 0:01 [du]
3505 root 0:00 [kworker/0:1]
3537 root 0:00 du -ch /usb/5big2_backup/2
3538 root 0:00 -sh
3550 root 0:00 ls --color -a /usb/5big2_backup/2
3551 root 0:00 [kworker/0:2]
3552 root 0:00 -sh
3560 root 0:00 ps
root@fvdw-sta-kirkwood:/ #

[Jocko]

You have at least 1 zombie process: du

So do

Code: Select all

kill -9 3537 3550

this should stop "du -ch /usb/5big2_backup/2" and "ls --color -a /usb/5big2_backup/2"

[imt17]

Well I also did a little looking online and ran TOP. I see that both the copy/backup process and the du -ch process to get the file size both show a status of "D". From what I saw on another forum is that the "D" is uninterruptible sleep and seems like its something like a frozen process. Hmmm. Here are the top results

Code: Select all

Mem: 351344K used, 160592K free, 0K shrd, 1163048K buff, 1163048K cached
CPU:  0.0% usr  0.0% sys  0.0% nic  0.0% idle  100% io  0.0% irq  0.0% sirq
Load average: 5.00 5.16 5.27 1/66 3624
  PID  PPID USER     STAT   VSZ %VSZ CPU %CPU COMMAND
 3624  3552 root     R     1228  0.2   0  0.0 top
 3142  1747 root     D     2132  0.4   0  0.0 /bin/ccp -r /md4/shares/2 /usb/5big2_backup
  694     1 root     S     1688  0.3   0  0.0 /sbin/discoverd
 3537     1 root     D     1256  0.2   0  0.0 du -ch /usb/5big2_backup/2
  705     1 root     S     1240  0.2   0  0.0 /sbin/telnetd
 3567   705 root     S     1236  0.2   0  0.0 -sh
 3538   705 root     S     1236  0.2   0  0.0 -sh
 3552   705 root     S     1236  0.2   0  0.0 -sh
    1     0 root     S     1232  0.2   0  0.0 /bin/sh
 1747   705 root     S     1232  0.2   0  0.0 -sh
  709     1 root     S      200  0.0   0  0.0 /sbin/buttons-nwsp2
 2120     2 root     SW       0  0.0   0  0.0 [md4_raid5]
  271     2 root     SW       0  0.0   0  0.0 [kswapd0]
 2934     2 root     DW       0  0.0   0  0.0 [usb-storage]
  414     2 root     DW       0  0.0   0  0.0 [kworker/u2:3]
    3     2 root     SW       0  0.0   0  0.0 [ksoftirqd/0]
 1803     2 root     SW<      0  0.0   0  0.0 [kworker/0:1H]
    7     2 root     SW       0  0.0   0  0.0 [rcu_preempt]
 2137     2 root     SW       0  0.0   0  0.0 [xfsaild/md4]

[fvdw]

Try

Code: Select all

killall -9 ccp

Or instead of ccp use /bin/ccp in the syntax above

If can not be stopped an it is doning nothing ignore it but that could be risky as we don't know were it is waiting for and may proceed again.
The killall command should blitz it down