Add a drive to RAID5

[RazzleUltra]

Right, this is really annoying me. I spent ages collating all my data onto my 3 disk n5200, and have realised that I haven't quite got enough room....grrrr

So I've got another HDD (exactly the same firmware as the other three, seagate 7200.10 750GB), and slotted it in. I've then tried to do a RAID5-RAID5 migration by selecting all the disks and choosing MIGRATE RAID.

This however just times out and IE brings back nothing.

Why the hell isn't this working? If I have to get everything back off the array and start again I'll be fuming....


NB It also suspends the SMB/CIFS service, and I have to reboot it to get it going again.

[giessle]

Hi,

please read my topic "RAID Migration - senseless ?".

Regards,
giessle

[RazzleUltra]

I had read it but I was hoping it was a different issue as you were going from RAID0 to RAID5 whereas I was just adding a drive to RAID5.

Oh well....

root@127.0.0.1:~# mdadm --detail /dev/md0
/dev/md0:
Version : 00.90.02
Creation Time : Wed Jan 31 05:34:20 2007
Raid Level : raid5
Array Size : 1463166976 (1395.38 GiB 1498.28 GB)
Device Size : 731583488 (697.69 GiB 749.14 GB)
Raid Devices : 3
Total Devices : 4
Preferred Minor : 0
Persistence : Superblock is persistent

Update Time : Wed Feb 7 04:58:39 2007
State : clean
Active Devices : 3
Working Devices : 4
Failed Devices : 0
Spare Devices : 1

Layout : left-symmetric
Chunk Size : 1024K

UUID : be37c47e:b8a59f08:091c21b6:7f42718f
Events : 0.74072

Number Major Minor RaidDevice State
0 8 2 0 active sync /dev/sda2
1 8 18 1 active sync /dev/sdb2
2 8 34 2 active sync /dev/sdc2

3 8 50 - spare /dev/sdd2

root@127.0.0.1:~# vgdisplay
--- Volume group ---
VG Name vg0
System ID
Format lvm2
Metadata Areas 1
Metadata Sequence No 3
VG Access read/write
VG Status resizable
MAX LV 0
Cur LV 2
Open LV 2
Max PV 0
Cur PV 1
Act PV 1
VG Size 1.36 TB
PE Size 4.00 MB
Total PE 357218
Alloc PE / Size 357069 / 1.36 TB
Free PE / Size 149 / 596.00 MB
VG UUID OxnxWz-GOyu-8k5s-gSE6-PQnN-Z1Qn-2YRAv4

root@127.0.0.1:~# pvdisplay
--- Physical volume ---
PV Name /dev/md0
VG Name vg0
PV Size 1.36 TB / not usable 0
Allocatable yes
PE Size (KByte) 4096
Total PE 357218
Free PE 149
Allocated PE 357069
PV UUID pyRcZz-PpAz-8wte-GyMo-qWrT-LJxC-1qz4dm

[RazzleUltra]

Can I not do something as simple as an MDADM --manage --add ?

If so, how? Don't realy want to bork my array by 'experimenting' on my own just yet.

l

[RazzleUltra]

Hmmm, found this snippet in the /img/bin/migrate_raid.sh file:

if [ -f "${old_raidconf}" ] ;then
if [ -f "${new_raidconf}" ] ;then
#stop service
sleep 2
/img/bin/service stop
sleep 1
#umount all file system
remove_allmount
#stop RAID
mdadm -S /dev/md0
#start migrate
#sleep 10
#exit
echo "Migrating RAID Starting ...." > /var/tmp/rss
echo "1" > /var/tmp/raidlock
#(/img/bin/raidreconf -o "$old_raidconf" -n "$new_raidconf" -r "$recoveryfile" -m "$mddisk" >/dev/null 2>&1 ;/img/bin/migrade_up.sh;)&
(sleep 1;/img/bin/raidreconf -o "$old_raidconf" -n "$new_raidconf" -m "$mddisk" >/dev/null 2>&1 ;sh -x /img/bin/resize_raid.sh 1 >/tmp/resize.tmp 2>&1;echo "0" > /var/tmp/raidlock)&
fi
fi

Shouldn't that last command not be remmed / #ed out? Doesn't this mean it doesn't actually call any commands to do the bloody migration?

[kevincy]

razzleultra said:

root@127.0.0.1:~# mdadm --detail /dev/md0

Could you type the follow command?
lvdisply
df
cat /proc/partition

[RazzleUltra]

login as: sys
sys@192.168.200.200's password:
root@127.0.0.1:~# lvdisplay
--- Logical volume ---
LV Name /dev/vg0/syslv
VG Name vg0
LV UUID Gc4DzH-uemY-1YDt-Rip2-RiAQ-aR70-ii7iVV
LV Write Access read/write
LV Status available
# open 1
LV Size 1.00 GB
Current LE 256
Segments 1
Allocation inherit
Read ahead sectors 0
Block device 253:0

--- Logical volume ---
LV Name /dev/vg0/lv0
VG Name vg0
LV UUID dRoUc3-F8x9-OcMb-HiK3-q1IG-elBB-7b9PWq
LV Write Access read/write
LV Status available
# open 1
LV Size 1.36 TB
Current LE 356813
Segments 1
Allocation inherit
Read ahead sectors 0
Block device 253:1

root@127.0.0.1:~# df
Filesystem 1k-blocks Used Available Use% Mounted on
/dev/ram0 39663 33673 5990 85% /
/dev/hda5 9523 2114 6917 23% /app
tmpfs 131072 376 130696 0% /tmpfs
/dev/hda3 30301 14562 14175 51% /cloopfs
/dev/cloop0 9911 6635 2764 71% /lib/modules
/dev/cloop1 39663 26610 11005 71% /opt
/dev/cloop2 3963 1857 1902 49% /img.SAVE/www
/dev/cloop3 3019 2407 459 84% /img.SAVE/bin
/dev/vg0/lv0 1415735580 1303467376 39192904 97% /raid/data
/dev/vg0/syslv 1015704 32872 930404 3% /raid/sys
root@127.0.0.1:~# cat /proc/partition
cat: /proc/partition: No such file or directory
root@127.0.0.1:~#

[Arctra]

kevincy said:

-snip-

cat /proc/partition

Don't you mean cat /proc/partitions?

[kevincy]

arbarbarb said:

kevincy said:

-snip-

cat /proc/partition

Don't you mean cat /proc/partitions?

sorry , you right!!

[kevincy]

razzleultra said:

login as: sys
sys@192.168.200.200's password:
root@127.0.0.1:~# lvdisplay
--- Logical volume ---
LV Name /dev/vg0/syslv
LV Size 1.00 GB

--- Logical volume ---
LV Name /dev/vg0/lv0
LV Size 1.36 TB

root@127.0.0.1:~# df
Filesystem 1k-blocks Used Available Use% Mounted on
//dev/vg0/lv0 1415735580 1303467376 39192904 97% /raid/data
/dev/vg0/syslv 1015704 32872 930404 3% /raid/sys

Hello razzleultra,
It seems ,your migration not success work .
Because your RAID mode just add a new spare disk .
My suggection is you are not success start RIAD Migration process .
So ,you can do migration again .

[RazzleUltra]

As below:

I'm still confused as to why it won't work. It seems to recognise the fourth drive perfectly well, but just doesn't correctly execute the migration.

login as: sys
sys@192.168.200.200's password:
root@127.0.0.1:~# cat /proc/partitions
major minor #blocks name

3 0 62592 hda
3 1 2032 hda1
3 2 11776 hda2
3 3 31296 hda3
3 4 1 hda4
3 5 9840 hda5
3 6 7600 hda6
240 0 10304 cloop0
240 1 41024 cloop1
240 2 4160 cloop2
240 3 3136 cloop3
8 0 732574584 sda
8 1 987966 sda1
8 2 731584035 sda2
8 16 732573527 sdb
8 17 987966 sdb1
8 18 731584035 sdb2
8 32 732574584 sdc
8 33 987966 sdc1
8 34 731584035 sdc2
8 48 732574584 sdd
8 49 987966 sdd1
8 50 731584035 sdd2
9 1 987904 md1
9 0 1463166976 md0
253 0 1048576 dm-0
253 1 1461506048 dm-1
root@127.0.0.1:~#

[kevincy]

razzleultra said:

As below:

I'm still confused as to why it won't work. It seems to recognise the fourth drive perfectly well, but just doesn't correctly execute the migration.

login as: sys
sys@192.168.200.200's password:
root@127.0.0.1:~# cat /proc/partitions
major minor #blocks name

3 0 62592 hda
3 1 2032 hda1
3 2 11776 hda2
3 3 31296 hda3
3 4 1 hda4
3 5 9840 hda5
3 6 7600 hda6
240 0 10304 cloop0
240 1 41024 cloop1
240 2 4160 cloop2
240 3 3136 cloop3
8 0 732574584 sda
8 1 987966 sda1
8 2 731584035 sda2
8 16 732573527 sdb
8 17 987966 sdb1
8 18 731584035 sdb2
8 32 732574584 sdc
8 33 987966 sdc1
8 34 731584035 sdc2
8 48 732574584 sdd
8 49 987966 sdd1
8 50 731584035 sdd2
9 1 987904 md1
9 0 1463166976 md0
253 0 1048576 dm-0
253 1 1461506048 dm-1
root@127.0.0.1:~#

It is more clear , problem maybe happen in UI or browser .
Before you start migration , maybe you can check UI whether have any javascript error .

[RazzleUltra]

I've just tried it again, and there are no errors showing on the interface. The Sun JVM doesn't start ever when I'm using the interface, and other than that I'm not sure where I would expect to see errors.

My exact procedure is:

1. Go to 'Migrate RAID'
2. Click on the 'RAID5 -> RAID5' radio button to ensure it's selected.
3. Put a tick in the 'Migrate Disk' column against all of the existing drives (the first three), and the 4th disk (which is a spare, but the same happened when it wasn't a spare).
4. Click 'Apply'
5. It shows on the browser status bar: "Waiting for https://<IP Address>/adm/setform.html... and sits there for about 10 seconds before going to an "Internet Explorer cannot display the webpage" site.
6. When I log back in and go to 'System', all the services have been stopped.

Aren't there a few commands I can run at the command prompt to expand the array manually, or will this break things as far as the interface is concerned?

[omega]

Hi razzleultra,

Kevin is talking about JavaScript errors. This is something totally different than Java (unfortunately both names are so similar). So JavaScript is done by the IE itself.

The services are stopped when doing the RAID migration with the command mdadm -S because the raidreconf program requires an unused raid device.

Kevin wants to know if there are any javascript errors because obviously the migration process is interrupted somehow. This is why you are getting the IE error and why the services are not started again.

So the main two questions are now:
1) how does it come that you additional drives has been configured as a spare drive?
2) How does it come that the PHP/javascript code terminates premature without giving at least an error message.

Sure Thecus support will investigate and tell you how to solve this issue....

Can I not do something as simple as an MDADM --manage --add ?

No, not with the used kernel on the N5200.
Starting with kernel 2.6.17 and mdadm 2.5.2 it will be possible as what I've read in the Internet. Maybe this link is interesting for you:
www.gagme.com/greg/linux/raid-lvm.php

Aren't there a few commands I can run at the command prompt to expand the array manually, or will this break things as far as the interface is concerned?

I don't know if you can but it is possible... ;D

Seriously, the Thecus software is doing a lot of checks before the "real" work and the "real" work is only issuing several commands. But I wouldn't take the risk.....
Even though I think I could do it myself I'm not experienced enough to tell you how to do it in a safe way, sorry.

Andreas

[RazzleUltra]

Thanks for the advice!

I manually added the drive as a spare after it failed to work with it as a non-spare.

I haven't seen any errors in the status bar of IE when it was running, so I'm guessing there's no errors that IE can see.

I've sent a message to Thecus and I'll see what they say...

EDIT: FYI I've tried to run the scripts under IE7 (Vista), IE7 (W2K3 Server) and Opera 8.5 (Vista) with the same result on all of them.