• We will be using a virtual machine in the faculty's cloud.
• When creating a virtual machine follow the steps in this tutorial.
• Create a VM
• When creating a virtual machine in the Launch Instance window:
  • For Availability zone, choose CAMPUS, CI or hp
  • Select Boot from image in Instance Boot Source section
  • Select SCGC Template v1 in Image Name section
  • Select a flavor that is at least m1.medium.
• The username for connecting to the VM is student
• Within the above virtual machine, we will be running three virtual machines (storage1, storage2, storage3)
• First, download the laboratory archive in the scgc/ directory:

  student@saisp:~$ cd scgc
  student@saisp:~/scgc$ wget --user=user-curs --ask-password http://repository.grid.pub.ro/cs/scgc/laboratoare/lab-03.zip
  student@saisp:~/scgc$ unzip lab-03.zip

• After unzipping the archive, several KVM image files (qcow2 format) should be present, as well as two scripts (lab03-start and lab03-stop)
• To run the virtual machines, use the lab03-start script:

  student@saisp:~/saisp$ ./lab03-start

• It may take a minute for the virtual machines to start
• In order to connect to each of the machines, use the following command (substitute X with 1, 2 or 3):

  student@saisp:~/saisp$ ssh -l root 192.168.1.X

• The password for both student and root users is student

Start by connecting to storage1:

student@saisp:~/saisp$ ssh -l root 192.168.1.1

mdadm is a tool used to manage RAID arrays.

root@storage1:~# apt-get update
[...]
root@storage1:~# apt-get install mdadm
[...]
[ ok ] Assembling MD array md0...done (started [3/3]).
[ ok ] Generating udev events for MD arrays...done.
update-rc.d: warning: start and stop actions are no longer supported; falling back to defaults
[ ok ] Starting MD monitoring service: mdadm --monitor.
Processing triggers for initramfs-tools (0.115) ...
update-initramfs: Generating /boot/initrd.img-3.12-1-amd64
root@storage1:~#

If prompted whether the root file system needs any arrays, leave the field blank (or enter none).

Notice that a RAID0 array has been configured during the installation process. Let us inspect it:

root@storage1:~# cat /proc/mdstat
Personalities : [raid0]
md0 : active raid0 sdb1[0] sdd1[2] sdc1[1]
      3139584 blocks super 1.2 512k chunks
 
unused devices: <none>

We can inspect the existing raid array in detail with the following command:

root@storage1:~# mdadm --detail /dev/md0
/dev/md0:
        Version : 1.2
  Creation Time : Mon Mar 17 00:49:44 2014
     Raid Level : raid0
     Array Size : 3139584 (2.99 GiB 3.21 GB)
   Raid Devices : 3
  Total Devices : 3
    Persistence : Superblock is persistent
 
    Update Time : Mon Mar 17 00:49:44 2014
          State : clean
 Active Devices : 3
Working Devices : 3
 Failed Devices : 0
  Spare Devices : 0
 
     Chunk Size : 512K
 
           Name : raid:0
           UUID : 7c853116:6277002c:1799d9e1:5a0eadcd
         Events : 0
 
    Number   Major   Minor   RaidDevice State
       0       8       17        0      active sync   /dev/sdb1
       1       8       33        1      active sync   /dev/sdc1
       2       8       49        2      active sync   /dev/sdd1

We will attempt to remove one of the disks from the array:

root@storage1:~# mdadm /dev/md0 --fail /dev/sdd1
mdadm: set device faulty failed for /dev/sdd1:  Device or resource busy

Let us remove the existing RAID0 array and replace it with a RAID1 array containing the same disks. We begin by stopping the exiting array.

root@storage1:~# mdadm --stop /dev/md0
mdadm: stopped /dev/md0
root@storage1:~# mdadm --detail /dev/md0
mdadm: cannot open /dev/md0: No such file or directory

We proceed to zero the superblocks of the previously used partition to clean them.

root@storage1:~# mdadm --zero-superblock /dev/sdb1
root@storage1:~# mdadm --zero-superblock /dev/sdc1
root@storage1:~# mdadm --zero-superblock /dev/sdd1

Finally, we can create the new array.

root@storage1:~# mdadm --create /dev/md0 --level=1 --raid-devices=3 /dev/sdb1 /dev/sdc1 /dev/sdd1
mdadm: Note: this array has metadata at the start and
    may not be suitable as a boot device.  If you plan to
    store '/boot' on this device please ensure that
    your boot-loader understands md/v1.x metadata, or use
    --metadata=0.90
Continue creating array? y
mdadm: Defaulting to version 1.2 metadata
 
mdadm: array /dev/md0 started.
root@storage1:~# mdadm --detail /dev/md0
/dev/md0:
        Version : 1.2
  Creation Time : Sat Mar 10 22:14:33 2018
     Raid Level : raid1
     Array Size : 1046976 (1022.61 MiB 1072.10 MB)
  Used Dev Size : 1046976 (1022.61 MiB 1072.10 MB)
   Raid Devices : 3
  Total Devices : 3
    Persistence : Superblock is persistent
 
    Update Time : Sat Mar 10 22:14:50 2018
          State : clean
 Active Devices : 3
Working Devices : 3
 Failed Devices : 0
  Spare Devices : 0
 
           Name : storage1:0  (local to host storage1)
           UUID : e1b180bf:b5232a89:d0420e5c:f9d74013
         Events : 17
 
    Number   Major   Minor   RaidDevice State
       0       8       17        0      active sync   /dev/sdb1
       1       8       33        1      active sync   /dev/sdc1
       2       8       49        2      active sync   /dev/sdd1

Let us remove a disk from the newly created array.

root@storage1:~# mdadm /dev/md0 --fail /dev/sdd1
mdadm: set /dev/sdd1 faulty in /dev/md0
root@storage1:~# mdadm --detail /dev/md0
/dev/md0:
[...]
    Number   Major   Minor   RaidDevice State
       0       8       17        0      active sync   /dev/sdb1
       1       8       33        1      active sync   /dev/sdc1
       4       0        0        4      removed
 
       2       8       49        -      faulty   /dev/sdd1
root@storage1:~# mdadm /dev/md0 --remove /dev/sdd1
mdadm: hot removed /dev/sdd1 from /dev/md0
root@storage1:~# mdadm --detail /dev/md0
/dev/md0:
[...]
    Number   Major   Minor   RaidDevice State
       0       8       17        0      active sync   /dev/sdb1
       1       8       33        1      active sync   /dev/sdc1
       4       0        0        4      removed

Using partitions sdb2, sdc2 and sdd2, create the md1 RAID5 array on storage1. Do the same setup on storage2. We will use it at a later stage.

The result should look like below

root@storage1:~# mdadm --detail /dev/md1
/dev/md1:
        Version : 1.2
  Creation Time : Sat Mar 10 22:31:42 2018
     Raid Level : raid5
     Array Size : 2095104 (2046.34 MiB 2145.39 MB)
  Used Dev Size : 1047552 (1023.17 MiB 1072.69 MB)
   Raid Devices : 3
  Total Devices : 3
    Persistence : Superblock is persistent
 
    Update Time : Sat Mar 10 22:31:51 2018
          State : clean, degraded, recovering
 Active Devices : 2
Working Devices : 3
 Failed Devices : 0
  Spare Devices : 1
 
         Layout : left-symmetric
     Chunk Size : 512K
 
 Rebuild Status : 59% complete
 
           Name : storage1:1  (local to host storage1)
           UUID : 129b8d62:eba119a1:172adf60:1adb20ee
         Events : 10
 
    Number   Major   Minor   RaidDevice State
       0       8       18        0      active sync   /dev/sdb2
       1       8       34        1      active sync   /dev/sdc2
       3       8       50        2      spare rebuilding   /dev/sdd2

Mark sdb2 as faulty and then remove it from the array.

root@storage1:~# mdadm --detail /dev/md1
/dev/md1:
        Version : 1.2
  Creation Time : Sat Mar 10 22:31:42 2018
     Raid Level : raid5
     Array Size : 2095104 (2046.34 MiB 2145.39 MB)
  Used Dev Size : 1047552 (1023.17 MiB 1072.69 MB)
   Raid Devices : 3
  Total Devices : 2
    Persistence : Superblock is persistent
 
    Update Time : Sat Mar 10 22:36:49 2018
          State : clean, degraded
 Active Devices : 2
Working Devices : 2
 Failed Devices : 0
  Spare Devices : 0
 
         Layout : left-symmetric
     Chunk Size : 512K
 
           Name : storage1:1  (local to host storage1)
           UUID : 129b8d62:eba119a1:172adf60:1adb20ee
         Events : 21
 
    Number   Major   Minor   RaidDevice State
       0       0        0        0      removed
       1       8       34        1      active sync   /dev/sdc2
       3       8       50        2      active sync   /dev/sdd2

We can add a disk in place of the one removed. In this case, we will use the same disk, sdb2.

root@storage1:~# mdadm /dev/md1 --add /dev/sdb2
mdadm: added /dev/sdb2
root@storage1:~# mdadm --detail /dev/md1
/dev/md1:
        Version : 1.2
  Creation Time : Sat Mar 10 22:31:42 2018
     Raid Level : raid5
     Array Size : 2095104 (2046.34 MiB 2145.39 MB)
  Used Dev Size : 1047552 (1023.17 MiB 1072.69 MB)
   Raid Devices : 3
  Total Devices : 3
    Persistence : Superblock is persistent
 
    Update Time : Sat Mar 10 22:37:09 2018
          State : clean, degraded, recovering
 Active Devices : 2
Working Devices : 3
 Failed Devices : 0
  Spare Devices : 1
 
         Layout : left-symmetric
     Chunk Size : 512K
 
 Rebuild Status : 25% complete
 
           Name : storage1:1  (local to host storage1)
           UUID : 129b8d62:eba119a1:172adf60:1adb20ee
         Events : 27
 
    Number   Major   Minor   RaidDevice State
       4       8       18        0      spare rebuilding   /dev/sdb2
       1       8       34        1      active sync   /dev/sdc2
       3       8       50        2      active sync   /dev/sdd2

In order to make our RAID configuration persistent, we can use the following command

root@storage1:~# mdadm --detail --scan >> /etc/mdadm/mdadm.conf

Reboot storage1 and verify that both md0 and md1 are active.

root@storage1:~# reboot

GlusterFS is a distributed file system used to aggregate storage from multiple sources. We will use it in conjunction with a RAID array to test various forms of redundancy.

XFS is the recommended file system for a GlusterFS configuration. Ext4 is also supported.

Begin by installing xfsprogs:

root@storage1:~# apt-get install xfsprogs
[...]
Setting up xfsprogs (3.2.1) ...
Processing triggers for libc-bin (2.17-97) ...
root@storage1:~#

Proceed by making a XFS partition. Use the previously created RAID5 array.

root@storage1:~# fdisk /dev/md1
Device contains neither a valid DOS partition table, nor Sun, SGI or OSF disklabel
Building a new DOS disklabel with disk identifier 0x95711919.
Changes will remain in memory only, until you decide to write them.
After that, of course, the previous content won't be recoverable.
 
Warning: invalid flag 0x0000 of partition table 4 will be corrected by w(rite)
 
Command (m for help): n
Partition type:
   p   primary (0 primary, 0 extended, 4 free)
   e   extended
Select (default p):
Using default response p
Partition number (1-4, default 1):
Using default value 1
First sector (2048-4190207, default 2048): 
Using default value 2048
Last sector, +sectors or +size{K,M,G} (2048-4190207, default 4190207):
Using default value 4190207
 
Command (m for help): q
 
root@storage1:~# mkfs.xfs -i size=512 /dev/md1
log stripe unit (524288 bytes) is too large (maximum is 256KiB)
log stripe unit adjusted to 32KiB
meta-data=/dev/md1               isize=512    agcount=8, agsize=65408 blks
         =                       sectsz=512   attr=2, projid32bit=1
         =                       crc=0        finobt=0
data     =                       bsize=4096   blocks=523264, imaxpct=25
         =                       sunit=128    swidth=256 blks
naming   =version 2              bsize=4096   ascii-ci=0 ftype=0
log      =internal log           bsize=4096   blocks=2560, version=2
         =                       sectsz=512   sunit=8 blks, lazy-count=1
realtime =none                   extsz=4096   blocks=0, rtextents=0

Mount the created partition on a local path.

root@storage1:~# mkdir /export
root@storage1:~# echo "/dev/md1 /export xfs defaults 1 2" >> /etc/fstab
root@storage1:~# mount /export
root@storage1:~# df -h | grep export
/dev/md1        2.0G   33M  2.0G   2% /export

Install glusterfs-server

root@storage1:~# apt-get install glusterfs-server
[...]
Setting up glusterfs-common (3.5.2-2+deb8u3) ...
Setting up glusterfs-client (3.5.2-2+deb8u3) ...
Setting up glusterfs-server (3.5.2-2+deb8u3) ...
[ ok ] Starting glusterd service: glusterd.
Setting up dmsetup (2:1.02.90-2.2+deb8u1) ...
update-initramfs: deferring update (trigger activated)
Processing triggers for libc-bin (2.17-97) ...
Processing triggers for initramfs-tools (0.115) ...
update-initramfs: Generating /boot/initrd.img-3.12-1-amd64
root@storage1:~#

Now, let us connect the two hosts.

root@storage1:~# gluster peer probe storage2
peer probe: success.
root@storage1:~# gluster peer status
Number of Peers: 1
 
Hostname: storage2
Uuid: 7faf0a96-48ea-4c23-91af-80311614fd57
State: Peer in Cluster (Connected)
root@storage1:~#

Once the hosts are connected, we can create a GlusterFS volume using partitions on both machines.

root@storage1:~# gluster volume create scgc transport tcp storage1:/export/brick1 storage2:/export/brick1
volume create: scgc: success: please start the volume to access data
root@storage1:~# gluster volume info
 
Volume Name: scgc
Type: Distribute
Volume ID: 91f6f6f3-9473-48e2-b49a-e8dcbe5e45e0
Status: Created
Number of Bricks: 2
Transport-type: tcp
Bricks:
Brick1: storage1:/export/brick1
Brick2: storage2:/export/brick1

We will setup access permission for our network and start the volume.

root@storage1:~# gluster volume set scgc auth.allow 192.168.1.*
volume set: success
root@storage1:~# gluster volume start scgc
volume start: scgc: success

We will now use storage3 as a GlusterFS client and mount the scgc volume.

root@storage3:~# apt-get install glusterfs-client
[...]
Setting up glusterfs-common (3.5.2-2+deb8u3) ...
Setting up glusterfs-client (3.5.2-2+deb8u3) ...
Setting up dmsetup (2:1.02.90-2.2+deb8u1) ...
update-initramfs: deferring update (trigger activated)
Processing triggers for libc-bin (2.17-97) ...
Processing triggers for initramfs-tools (0.115) ...
update-initramfs: Generating /boot/initrd.img-3.12-1-amd64
root@storage3:~# mkdir /export
root@storage3:~# mount -t glusterfs storage1:/scgc /export
root@storage3:~# df -h | grep export
storage1:/scgc  4.0G   66M  4.0G   2% /export

Test that removing and adding volumes from the RAID arrays does not affect the GlusterFS volume. What limitations can you notice?

In the above setup, we used RAID5 in order to provide redundancy and GlusterFS to aggregate storage.

Now use GlusterFS to replicate data. Use the same disks on the two hosts as previously (sdb2, sdc2, sdd2) and arrange them in a suitable RAID array.

To remove the existing GlusterFS volume, do the following:

root@storage3:~# umount /export
 
root@storage1:~# gluster volume stop scgc
Stopping volume will make its data inaccessible. Do you want to continue? (y/n) y
root@storage1:~# gluster volume delete scgc
Deleting volume will erase all information about the volume. Do you want to continue? (y/n) y
volume delete: scgc: success
root@storage1:~# rm -rf /export/brick1/
 
root@storage2:~# rm -rf /export/brick1/