J'ai créé une matrice RAID10 en utilisant 4 disques 75G, pour créer un stockage de 150G.
Une fois tout terminé (y compris la synchronisation initiale), tout semblait bien, sauf la sortie df -h
qui ne montrait qu'un stockage 73G sur le point de montage désigné.
Détails:
- La machine est une instance Ubuntu 11.10 m1.large sur Amazon EC2.
- Les 4 disques sont des disques EBS, chacun a une taille de 75G.
- La matrice RAID10 a été créée à l'aide du script suivant:
-
#!/bin/sh
disk1="/dev/sdh1"
disk2="/dev/sdh2"
disk3="/dev/sdh3"
disk4="/dev/sdh4"
echo "*** Verifying existence of 4 volumes $disk1, $disk2, $disk3 and $disk4"
if [ -b "$disk1" -a -b "$disk2" -a -b "$disk3" -a -b "$disk4" ]; then
echo "# Found expected block devices."
else
echo "!!! Did not find expected block devices. Error."
exit -1
fi
until read -p "??? - How big (in GB) are the disks (They should be the same size)? " disk_size && [ $disk_size ]; do
echo "Please enter a disk size."
done
lv_size=$(echo "scale=2; $disk_size * 2.0" | bc)
echo "*** Assuming a per disk size of $disk_size gigs, will create a logical volume of $lv_size gigs, with $lv_size reserved for snapshots"
echo "*** Partitioning disks..."
echo "~ Partitioning $disk1"
echo ',,L' | sfdisk $disk1
echo "~ Partitioning $disk2"
echo ',,L' | sfdisk $disk2
echo "~ Partitioning $disk3"
echo ',,L' | sfdisk $disk3
echo "~ Partitioning $disk4"
echo ',,L' | sfdisk $disk4
sleep 6
echo "*** Creating /dev/md0 as a RAID 10"
/sbin/mdadm /dev/md0 --create --level=10 --raid-devices=4 $disk1 $disk2 $disk3 $disk4
echo " ~ Allocating /dev/md0 as a physical volume."
/sbin/pvcreate /dev/md0
echo " ~ Allocating a Volume Group 'mongodb_vg'"
/sbin/vgcreate -s 64M mongodb_vg /dev/md0
echo " ~ Creating a Logical Volume 'mongodb_lv'"
num_extents=$(echo "$disk_size * 1000 / 64" | bc)
/sbin/lvcreate -l $num_extents -nmongodb_lv mongodb_vg
echo " ~ Formatting the new volume (/dev/mongodb_vg/mongodb_lv) with EXT4"
/sbin/mkfs.ext4 /dev/mongodb_vg/mongodb_lv
echo " ~ Done! Go ahead and mount the new filesystem. Suggested FStab: "
echo " /dev/mongodb_vg/mongodb_lv /data ext4 defaults,noatime 0 0"
Voici la sortie que j'ai obtenue:
*** Verifying existence of 4 volumes /dev/xvdh1, /dev/xvdh2, /dev/xvdh3 and /dev/xvdh4
# Found expected block devices.
??? - How big (in GB) are the disks (They should be the same size)? 75
*** Assuming a per disk size of 75 gigs, will create a logical volume of 150.0 gigs, with 150.0 reserved for snapshots
*** Partitioning disks...
~ Partitioning /dev/xvdh1
Checking that no-one is using this disk right now ...
BLKRRPART: Invalid argument
OK
Disk /dev/xvdh1: 9790 cylinders, 255 heads, 63 sectors/track
sfdisk: ERROR: sector 0 does not have an msdos signature
/dev/xvdh1: unrecognized partition table type
Old situation:
No partitions found
New situation:
Units = cylinders of 8225280 bytes, blocks of 1024 bytes, counting from 0
Device Boot Start End #cyls #blocks Id System
/dev/xvdh1p1 0+ 9789 9790- 78638174+ 83 Linux
/dev/xvdh1p2 0 - 0 0 0 Empty
/dev/xvdh1p3 0 - 0 0 0 Empty
/dev/xvdh1p4 0 - 0 0 0 Empty
Warning: no primary partition is marked bootable (active)
This does not matter for LILO, but the DOS MBR will not boot this disk.
Successfully wrote the new partition table
Re-reading the partition table ...
BLKRRPART: Invalid argument
If you created or changed a DOS partition, /dev/foo7, say, then use dd(1)
to zero the first 512 bytes: dd if=/dev/zero of=/dev/foo7 bs=512 count=1
(See fdisk(8).)
~ Partitioning /dev/xvdh2
Checking that no-one is using this disk right now ...
BLKRRPART: Invalid argument
OK
Disk /dev/xvdh2: 9790 cylinders, 255 heads, 63 sectors/track
sfdisk: ERROR: sector 0 does not have an msdos signature
/dev/xvdh2: unrecognized partition table type
Old situation:
No partitions found
New situation:
Units = cylinders of 8225280 bytes, blocks of 1024 bytes, counting from 0
Device Boot Start End #cyls #blocks Id System
/dev/xvdh2p1 0+ 9789 9790- 78638174+ 83 Linux
/dev/xvdh2p2 0 - 0 0 0 Empty
/dev/xvdh2p3 0 - 0 0 0 Empty
/dev/xvdh2p4 0 - 0 0 0 Empty
Warning: no primary partition is marked bootable (active)
This does not matter for LILO, but the DOS MBR will not boot this disk.
Successfully wrote the new partition table
Re-reading the partition table ...
BLKRRPART: Invalid argument
If you created or changed a DOS partition, /dev/foo7, say, then use dd(1)
to zero the first 512 bytes: dd if=/dev/zero of=/dev/foo7 bs=512 count=1
(See fdisk(8).)
~ Partitioning /dev/xvdh3
Checking that no-one is using this disk right now ...
BLKRRPART: Invalid argument
OK
Disk /dev/xvdh3: 9790 cylinders, 255 heads, 63 sectors/track
sfdisk: ERROR: sector 0 does not have an msdos signature
/dev/xvdh3: unrecognized partition table type
Old situation:
No partitions found
New situation:
Units = cylinders of 8225280 bytes, blocks of 1024 bytes, counting from 0
Device Boot Start End #cyls #blocks Id System
/dev/xvdh3p1 0+ 9789 9790- 78638174+ 83 Linux
/dev/xvdh3p2 0 - 0 0 0 Empty
/dev/xvdh3p3 0 - 0 0 0 Empty
/dev/xvdh3p4 0 - 0 0 0 Empty
Warning: no primary partition is marked bootable (active)
This does not matter for LILO, but the DOS MBR will not boot this disk.
Successfully wrote the new partition table
Re-reading the partition table ...
BLKRRPART: Invalid argument
If you created or changed a DOS partition, /dev/foo7, say, then use dd(1)
to zero the first 512 bytes: dd if=/dev/zero of=/dev/foo7 bs=512 count=1
(See fdisk(8).)
~ Partitioning /dev/xvdh4
Checking that no-one is using this disk right now ...
BLKRRPART: Invalid argument
OK
Disk /dev/xvdh4: 9790 cylinders, 255 heads, 63 sectors/track
sfdisk: ERROR: sector 0 does not have an msdos signature
/dev/xvdh4: unrecognized partition table type
Old situation:
No partitions found
New situation:
Units = cylinders of 8225280 bytes, blocks of 1024 bytes, counting from 0
Device Boot Start End #cyls #blocks Id System
/dev/xvdh4p1 0+ 9789 9790- 78638174+ 83 Linux
/dev/xvdh4p2 0 - 0 0 0 Empty
/dev/xvdh4p3 0 - 0 0 0 Empty
/dev/xvdh4p4 0 - 0 0 0 Empty
Warning: no primary partition is marked bootable (active)
This does not matter for LILO, but the DOS MBR will not boot this disk.
Successfully wrote the new partition table
Re-reading the partition table ...
BLKRRPART: Invalid argument
If you created or changed a DOS partition, /dev/foo7, say, then use dd(1)
to zero the first 512 bytes: dd if=/dev/zero of=/dev/foo7 bs=512 count=1
(See fdisk(8).)
*** Creating /dev/md0 as a RAID 10
mdadm: partition table exists on /dev/xvdh1 but will be lost or
meaningless after creating array
mdadm: partition table exists on /dev/xvdh2 but will be lost or
meaningless after creating array
mdadm: partition table exists on /dev/xvdh3 but will be lost or
meaningless after creating array
mdadm: partition table exists on /dev/xvdh4 but will be lost or
meaningless after creating array
Continue creating array? y
mdadm: Defaulting to version 1.2 metadata
mdadm: array /dev/md0 started.
~ Allocating /dev/md0 as a physical volume.
Physical volume "/dev/md0" successfully created
~ Allocating a Volume Group 'mongodb_vg'
Volume group "mongodb_vg" successfully created
~ Creating a Logical Volume 'mongodb_lv'
Logical volume "mongodb_lv" created
~ Formatting the new volume (/dev/mongodb_vg/mongodb_lv) with EXT4
mke2fs 1.41.14 (22-Dec-2010)
Filesystem label=
OS type: Linux
Block size=4096 (log=2)
Fragment size=4096 (log=2)
Stride=128 blocks, Stripe width=256 blocks
4800512 inodes, 19185664 blocks
959283 blocks (5.00%) reserved for the super user
First data block=0
Maximum filesystem blocks=4294967296
586 block groups
32768 blocks per group, 32768 fragments per group
8192 inodes per group
Superblock backups stored on blocks:
32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 2654208,
4096000, 7962624, 11239424
Writing inode tables: done
Creating journal (32768 blocks): done
Writing superblocks and filesystem accounting information: done
This filesystem will be automatically checked every 36 mounts or
180 days, whichever comes first. Use tune2fs -c or -i to override.
~ Done! Go ahead and mount the new filesystem. Suggested FStab:
/dev/mongodb_vg/mongodb_lv /data ext4 defaults,noatime 0 0
Il s'agit de la sortie pertinente de df -h
:
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/mongodb_vg-mongodb_lv
73G 180M 69G 1% /ebsRaid
Ceci est la sortie de mdadm --detail /dev/md0
/dev/md0:
Version : 1.2
Creation Time : Wed Feb 29 10:14:39 2012
Raid Level : raid10
Array Size : 157283328 (150.00 GiB 161.06 GB)
Used Dev Size : 78641664 (75.00 GiB 80.53 GB)
Raid Devices : 4
Total Devices : 4
Persistence : Superblock is persistent
Update Time : Wed Feb 29 13:21:49 2012
State : clean
Active Devices : 4
Working Devices : 4
Failed Devices : 0
Spare Devices : 0
Layout : near=2
Chunk Size : 512K
Name : my.site.com:0 (local to host my.site.com)
UUID : CENSORED
Events : 19
Number Major Minor RaidDevice State
0 202 113 0 active sync /dev/xvdh1
1 202 114 1 active sync /dev/xvdh2
2 202 115 2 active sync /dev/xvdh3
3 202 116 3 active sync /dev/xvdh4
C'est la sortie de cat /proc/mdstat
:
Personalities : [raid10]
md0 : active raid10 xvdh4[3] xvdh3[2] xvdh2[1] xvdh1[0]
157283328 blocks super 1.2 512K chunks 2 near-copies [4/4] [UUUU]
unused devices: <none>
EDIT1:
Voici la sortie de lvdisplay -m:
--- Logical volume ---
LV Name /dev/mongodb_vg/mongodb_lv
VG Name mongodb_vg
LV UUID SEpGth-cXd3-ZFhy-XLHo-T5pV-gEd1-Tgancs
LV Write Access read/write
LV Status available
# open 1
LV Size 73.19 GiB
Current LE 1171
Segments 1
Allocation inherit
Read ahead sectors auto
- currently set to 4096
Block device 252:0
--- Segments ---
Logical extent 0 to 1170:
Type linear
Physical volume /dev/md0
Physical extents 0 to 1170
EDIT 2:
C'est la sortie de vgdisplay
:
--- Volume group ---
VG Name mongodb_vg
System ID
Format lvm2
Metadata Areas 1
Metadata Sequence No 2
VG Access read/write
VG Status resizable
MAX LV 0
Cur LV 1
Open LV 1
Max PV 0
Cur PV 1
Act PV 1
VG Size 149.94 GiB
PE Size 64.00 MiB
Total PE 2399
Alloc PE / Size 1171 / 73.19 GiB
Free PE / Size 1228 / 76.75 GiB
VG UUID CENSORED
lvdisply -m
et poster s'il vous plaît.Réponses:
Votre groupe de volumes n'utilise pas l'intégralité des extensions créées pour lui:
Vous pouvez ajouter des extensions supplémentaires à l'aide de la commande suivante:
VEUILLEZ LIRE LA PAGE DE MAN AVANT DE TAPER CECI DANS
Cette commande étendrait le groupe de volumes pour utiliser tous les extensions GRATUITES restantes (vous pouvez également sélectionner moins si vous souhaitez conserver certaines extensions libres). Il utilisera les extensions sur md0 pour ce faire.
Vous pouvez ensuite redimensionner la partition en ligne en utilisant:
Et cela devrait dire qu'il redimensionne en ligne. Je crois que cela résoudra votre problème, mais veuillez lire les pages de manuel et comprendre ce qu'elles font avant d'essayer. Je ne suis pas responsable de la destruction de vos disques.
En passant, RAID sur un EBS puis lvm sur le dessus semble être un niveau inutile de virtualisation des disques. Vous n'améliorerez pas les performances ni la sécurité des données en ajoutant du RAID supplémentaire. LVM fait déjà la mise en miroir / segmentation si je me souviens bien. Bien que vous puissiez techniquement exécuter LVM sur RAID sur LVM sur RAID à l'infini, je ne suis pas sûr que vous gagniez beaucoup en le faisant (même si je suis plus qu'heureux d'être souligné que j'ai tort sur ce point).
la source