https://wiki.ixheim.de/index.php?action=history&feed=atom&title=CEPH_Manuell 2026-06-29T06:36:25Z Versionsgeschichte dieser Seite in Xinux Wiki MediaWiki 1.35.1 https://wiki.ixheim.de/index.php?title=CEPH_Manuell&diff=50710&oldid=prev 2023-12-06T06:49:10Z

<p>Die Seite wurde neu angelegt: „ =Aufbau= ==Storage-Node und Monitor Nodes== *omar *nelson *bay ==Admin-Node== *kekistan =Vorbreitung= ==Installation von ssh und ntp== Auf allen Nodes *sudo…“</p> <p><b>Neue Seite</b></p><div><br /> =Aufbau=<br /> ==Storage-Node und Monitor Nodes==<br /> *omar<br /> *nelson<br /> *bay<br /> <br /> ==Admin-Node==<br /> *kekistan<br /> <br /> =Vorbreitung=<br /> ==Installation von ssh und ntp==<br /> Auf allen Nodes<br /> *sudo apt -y install ntp ssh<br /> *wget -q -O- 'https://download.ceph.com/keys/release.asc' | apt-key add -<br /> *echo deb http://download.ceph.com/debian-luminous/ $(lsb_release -sc) main | tee /etc/apt/sources.list.d/ceph.list<br /> *apt-get update<br /> <br /> ==ssh-key auf allen nodes==<br /> ===ssh-keygen===<br /> *root@kekistand:~# ssh-keygen <br /> ===ssh-copy-id===<br /> *root@kekistand:~# ssh-copy-id nelson<br /> *root@kekistand:~# ssh-copy-id bay<br /> *root@kekistand:~# ssh-copy-id omar<br /> ===Hilfskript===<br /> *cat /usr/local/sbin/root-all-nodes<br /> &lt;pre&gt;<br /> #!/bin/bash<br /> echo localhost<br /> $*<br /> LIST=&quot;bay omar nelson&quot;<br /> for k in $LIST<br /> do<br /> echo $k<br /> ssh $k $*<br /> done<br /> &lt;/pre&gt;<br /> <br /> ===Anlegen des Users cephadmin und keyhandling===<br /> *root-all-nodes useradd -s /bin/bash -m cephadmin<br /> *sudo -u cephadmin ssh-keygen -q<br /> *root-all-nodes mkdir /home/cephadmin/.ssh<br /> *root-all-nodes chown cephadmin.cephadmin /home/cephadmin/.ssh<br /> *root-all-nodes chmod 700 /home/cephadmin/.ssh<br /> *root-all-nodes 'echo &quot;cephadmin ALL=(ALL) NOPASSWD: ALL&quot; &gt; /etc/sudoers.d/ceph'<br /> *scp /home/cephadmin/.ssh/id_rsa.pub /tmp<br /> *scp /home/cephadmin/.ssh/id_rsa.pub bay:/tmp<br /> *scp /home/cephadmin/.ssh/id_rsa.pub nelson:/tmp<br /> *scp /home/cephadmin/.ssh/id_rsa.pub omar:/tmp<br /> *root-all-nodes cat /tmp/id_rsa.pub &gt; /home/cephadmin/.ssh/authorized_keys<br /> <br /> ==Purge==<br /> <br /> ==Install==<br /> *apt install ceph-deploy<br /> *su - cephadmin<br /> *mkdir my-first-ceph-cluster<br /> *cd my-first-ceph-cluster<br /> ==Purge==<br /> *ceph-deploy purge kekistan bay nelson omar<br /> *ceph-deploy purgedata kekistan bay nelson omar<br /> *ceph-deploy forgetkeys<br /> <br /> ==Monitor Nodes==<br /> *ceph-deploy new bay nelson omar<br /> *echo &quot;public network = 10.0.0.0/8&quot; &gt;&gt; ceph.conf<br /> *echo &quot;osd pool default size = 2&quot; &gt;&gt; ceph.conf<br /> <br /> ==Install Ceph==<br /> *ceph-deploy install kekistan bay nelson omar<br /> ==Initialisierung der Monitore und sammeln der der Keys==<br /> *ceph-deploy mon create-initial<br /> <br /> ==Vorbereiten der Storages==<br /> *ceph-deploy osd prepare bay:/dev/sdb nelson:/dev/sdb omar:/dev/sdb<br /> <br /> ==Aktivieren der Storages==<br /> *ceph-deploy osd activate bay:/dev/sdb1 nelson:/dev/sdb1 omar:/dev/sdb1<br /> ==Löschen eines Storages==<br /> Auf dem OSD<br /> *NR=1<br /> *ceph osd out $NR <br /> *sudo systemctl stop ceph-osd@$NR.service<br /> *ceph osd crush remove osd.$NR<br /> *ceph auth del osd.$NR<br /> *ceph osd rm $NR<br /> *umount /dev/sda1<br /> *parted -s /dev/sda rm 2<br /> *parted -s /dev/sda rm 1<br /> *partprobe<br /> <br /> ==Kopieren der Admin-keyrings==<br /> *ceph-deploy admin kekistan bay nelson omar<br /> *sudo chmod 644 /etc/ceph/ceph.client.admin.keyring<br /> <br /> ==Status check==<br /> *ceph osd tree<br /> ID WEIGHT TYPE NAME UP/DOWN REWEIGHT PRIMARY-AFFINITY <br /> -1 1.44955 root default <br /> -2 0.48318 host bay <br /> 0 0.48318 osd.0 up 1.00000 1.00000 <br /> -3 0.48318 host nelson <br /> 1 0.48318 osd.1 up 1.00000 1.00000 <br /> -4 0.48318 host omar <br /> 2 0.48318 osd.2 up 1.00000 1.00000 <br /> *ceph health<br /> HEALTH_OK<br /> =Block Device Client=<br /> auf dem Client cephadmin einrichten und keys von admin node kopieren ... <br /> *ceph-deploy install bajor<br /> *ceph-deploy admin bajor<br /> *rbd create bar --size 4096 --image-feature layering<br /> *sudo rbd map bar --name client.admin<br /> *sudo mkfs.ext4 /dev/rbd0<br /> *sudo mkdir /mnt/ceph-block-device<br /> *sudo mount /dev/rbd0 /mnt/ceph-block-device<br /> <br /> =Anlegen eines Metadata Server(Voraussetzung für das CephFS)=<br /> *ceph-deploy mds create bay nelson omar<br /> *ceph osd pool create cephfs_data 1<br /> *ceph osd pool create cephfs_metadata 1<br /> *ceph fs new cephfs cephfs_metadata cephfs_data<br /> *ceph mds stat<br /> <br /> <br /> =Client File System=<br /> *installieren von seph-fs-common<br /> apt install ceph-fs-common<br /> *auf dem Client das secret speichern <br /> mkdir /etc/ceph/<br /> vi /etc/ceph/admin.secret<br /> <br /> <br /> *TIPP: secret auf dem client soll in der form sein<br /> *grep key ceph.client.admin.keyring | cut -f 2 -d =<br /> <br /> ÖLUJmb5rfHOH8GigijgGTfu==<br /> *eintrag in der fstab anlegen <br /> vi /etc/fstab<br /> 10.1.1.1,10.1.1.2,10.1.1.3:/ /mnt/cephfs ceph name=admin,secretfile=/etc/ceph/admin.secret 0 0<br /> *mountpunkt erstellen &amp; Files System mounten<br /> mkdir /mnt/cephfs<br /> mount /mnt/cephfs<br /> <br /> =Links=<br /> *http://docs.ceph.com/docs/master/start/<br /> *https://linoxide.com/ubuntu-how-to/create-ceph-cluster-ubuntu-16-04/<br /> *https://www.howtoforge.com/tutorial/how-to-install-a-ceph-cluster-on-ubuntu-16-04/<br /> *https://www.howtoforge.com/tutorial/how-to-mount-cephfs-on-centos-7/<br /> *http://www.admin-magazin.de/Das-Heft/2014/04/Die-verteilten-Speichersysteme-GlusterFS-und-Ceph-im-Vergleich/(offset)/12<br /> *https://www.sebastien-han.fr/blog/2012/07/06/nfs-over-rbd/</div>

Thomas.will