software-defined-storage

Benvenuti al corso su Software defined storage!

Parte 1

slides

Parte 2

dashboard

    -        Lanciare 4 template -> Images - shared with me - count 4
    -        Creare 2 volumi STD 50GB
    -        Rinominare le VM da Horizon come admin,mon1,osd1,osd2
    -        Dare IP pubblico ad admin
    -        Dentro a admin
                    -        hostname admin
                    -        echo "admin" > /etc/hostname
                    -        ssh-keygen

sistemare hosts - prendere ip reali da dashboard

echo "192.168.0.x mon1" >> /etc/hosts echo "192.168.0.x osd1" >> /etc/hosts echo "192.168.0.x osd2" >> /etc/hosts echo "192.168.0.x admin" >> /etc/hosts

copiare chiavi

ssh-keygen ssh-copy-id root@mon1 ssh-copy-id root@osd1 ssh-copy-id root@osd2

settare hostname

ssh mon1 'hostname mon1' ssh mon1 'echo "mon1" > /etc/hostname'
ssh osd1 'hostname osd1' ssh osd1 'echo "osd1" > /etc/hostname' ssh osd2 'hostname osd2' ssh osd2 'echo "osd2" > /etc/hostname'

copiare hosts

scp /etc/hosts root@mon1:/etc/hosts scp /etc/hosts root@osd1:/etc/hosts scp /etc/hosts root@osd2:/etc/hosts

installare pacchetti

apt update apt install -y ceph-deploy ntp ssh mon1 'apt install -y python ntp' ssh osd1 'apt install -y python ntp' ssh osd2 'apt install -y python ntp'

installare ceph

mkdir loged-cluster cd loged-cluster ceph-deploy new mon1 echo "osd pool default size = 2" >> ceph.conf ceph-deploy install admin mon1 osd1 osd2 ceph-deploy mon create-initial ceph-deploy admin admin mon1 osd1 osd2 ceph-deploy osd prepare osd1:/dev/vdb osd2:/dev/vdb ceph -w ceph -s

espansione cluster HA

lanciare 3 template - rinominarne 2 mon2,mon3 echo "192.168.0.x mon2" >> /etc/hosts echo "192.168.0.x mon3" >> /etc/hosts ssh-copy-id root@mon2 ssh-copy-id root@mon3 ssh mon2 'hostname mon2' ssh mon2 'echo "mon2" > /etc/hostname'
ssh mon2 reboot ssh mon3 'hostname mon3' ssh mon3 'echo "mon3" > /etc/hostname'
ssh mon3 reboot scp /etc/hosts root@mon1:/etc/hosts scp /etc/hosts root@mon2:/etc/hosts scp /etc/hosts root@mon3:/etc/hosts scp /etc/hosts root@osd1:/etc/hosts scp /etc/hosts root@osd2:/etc/hosts ssh mon2 'apt install -y python ntp' ssh mon3 'apt install -y python ntp' ceph-deploy install mon2 mon3 ceph-deploy mon add mon2 ceph-deploy mon add mon3 ceph -s ceph mon dump

rinominare terzo template osd3

echo "192.168.0.x osd3" >> /etc/hosts ssh-copy-id root@osd3 ssh osd3 'hostname osd3' ssh osd3 'echo "osd3" > /etc/hostname'
ssh osd3 reboot scp /etc/hosts root@mon1:/etc/hosts scp /etc/hosts root@mon2:/etc/hosts scp /etc/hosts root@mon3:/etc/hosts scp /etc/hosts root@osd1:/etc/hosts scp /etc/hosts root@osd2:/etc/hosts scp /etc/hosts root@osd3:/etc/hosts ssh osd3 'apt install -y python ntp' ceph-deploy install osd3 ceph-deploy osd prepare osd3:/dev/vdb ceph osd tree ceph osd dump ceph osd pool ls ceph osd pool get rbd min_size ceph osd pool get rbd size ceph -w

aggangiare volume a VM

rbd -p rbd create loged --size 10G --image-feature layering rbd -p rbd info loged rbd map -p rbd loged rbd showmapped fdisk /dev/rbd0 mkfs.xfs /dev/rbd0p1 mount /dev/rbd0p1 /mnt cd /mnt touch loged

testare ridondanza

ssh osd3 shutdown -h now ceph -w ssh mon3 shutdown -h now ceph -w ceph -s

Parte 3

slides

dashboard

    -        Lanciare 2 istanze e3standard.x4 -> Images ubuntu 16.04
    -        Creare 2 volumi STD 50GB
    -        Rinominare le VM da Horizon come glusterfs-1 e glusterfs-2

sistemare hosts - prendere ip reali da dashboard

echo "192.168.0.x glusterfs-1" >> /etc/hosts echo "192.168.0.x glusterfs-2" >> /etc/hosts

copiare chiavi

ssh-keygen ssh-copy-id root@glusterfs-1 ssh-copy-id root@glusterfs-2

settare hostname

ssh glusterfs-1 'hostname glusterfs-1' ssh glusterfs-1 'echo "glusterfs-1" > /etc/hostname' ssh glusterfs-1 'hostname glusterfs-2' ssh glusterfs-1 'echo "glusterfs-2" > /etc/hostname'

copiare hosts

scp /etc/hosts root@glusterfs-2:/etc/hosts

preparare dischi

parted -s /dev/vdb mklabel gpt parted --align optimal -s /dev/vdb mkpart primary xfs 0% 100% mkfs.xfs /dev/vdb1 mkdir /mnt/storage-brick mount /dev/vdb1 /mnt/storage-brick

install

apt install glusterfs-server -y

peer

gluster peer probe glusterfs-1 gluster peer probe glusterfs-2

create volume

gluster volume create volume0 replica 2 transport tcp glusterfs-1:/mnt/storage-brick/volume0 glusterfs-2:/mnt/storage-brick/volume0

opzioni mount da fstab

glusterfs-1:/volume0 /mnt/cloudstorage glusterfs acl,backupvolfile-server=glusterfs-2,fetch-attempts=2,log-level=WARNING,log-file=/var/log/gluster.log 0 0