Ceci est une ancienne révision du document !
Table des matières
Dernière mise-à-jour : 2021/01/26 09:42
HAR200 - Gestion de la Haute Disponibilité avec OpenSVC sous RHEL 7
Contenud du Module
- HAR200 - Gestion de la Haute Disponibilité avec OpenSVC sous CentOS 7
- Contenu du Module
- Présentation
- L'Agent
- Collector
- Mise en Place
- Préparation des Machines Virtuelles
- Configuration du node1.opensvc.loc
- Configuration du node2.opensvc.loc
- Mise en place des LVM de test
- Configuration du node1.opensvc.loc ET node2.opensvc.loc
- Installation d'opensvc-agent
- Installation sur node1.opensvc.loc et node2.opensvc.loc
- Clefs SSH
- Création d'un Service
- Créer le fichier de service
- Le Répertoire des Scripts de Démarrage des Services
- Service Management Facility
- Intégrer une Application
- Applications launcher directory
- Obtenir le binaire de l'application
- Service Failover
- Haute Disponibilité
Présentation
Le projet OpenSVC est né au mois de novembre 2009. OpenSVC comprend deux composants logiciels :
L'Agent
L'agent OpenSVC est un Gestionnaire de Ressources du Cluster publié sous la licence GNU General Public License v2 ayant des fonctions de reporting et de gestion de la configuration. Couplé avec un logiciel heartbeat, l'agent est un outil de gestion d'un cluster complet.
Plus d'information concernent l'agent est disponible ici.
Collector
Le Collector est un logiciel commercial optionnel, compatible web 2.0. Comme son nom indique, il collecte des informations à partir des agents ainsi qu'à partir des éléments de l'infrastructure du site tel le réseau, les SANs, les serveurs etc.)
Mise en Place
Préparation des Machines Virtuelles
A partir de votre machine virtuelle CentOS 7, créez 2 clones complets :
VBoxManage clonevm CentOS_7 --name="node1.i2tch.loc" --register --mode=all VBoxManage clonevm CentOS_7 --name="node2.i2tch.loc" --register --mode=all
Modifiez la configuration réseau des deux clones :
| Adaptateur | Carte 1 | Carte 2 | Carte 3 |
|---|---|---|---|
| Type de réseau | NAT | intnet | intnet |
Important - Dans Virtual Box > Paramètres de node2.opensvc.loc > Réseau > Carte 1 > Redirection de ports, Modifiez le port hôte ssh en 4022.
Démarrez les machines virtuelles node1.opensvc.loc et node2.opensvc.loc et modifiez les noms d'hôtes ainsi :
[root@centos7 ~]# nmcli general hostname node1.opensvc.loc [root@centos7 ~]# hostname node1.opensvc.loc
[root@centos7 ~]# nmcli general hostname node2.opensvc.loc [root@centos7 ~]# hostname node2.opensvc.loc
Important - Déconnectez-vous de et re-connectez-vous à chaque VM.
Vérifiez la configuration réseau sur chaque noeud :
[root@node1 ~]# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN qlen 1
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: enp0s3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 08:00:27:35:2a:bc brd ff:ff:ff:ff:ff:ff
inet 10.0.2.15/24 brd 10.0.2.255 scope global dynamic enp0s3
valid_lft 86321sec preferred_lft 86321sec
inet6 fe80::9e30:f271:f425:faa2/64 scope link
valid_lft forever preferred_lft forever
3: enp0s8: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 08:00:27:9a:f3:96 brd ff:ff:ff:ff:ff:ff
inet6 fe80::c141:816b:e364:4dc/64 scope link
valid_lft forever preferred_lft forever
4: enp0s9: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 08:00:27:84:d3:59 brd ff:ff:ff:ff:ff:ff
inet6 fe80::e1a1:e92e:b0ab:3b40/64 scope link
valid_lft forever preferred_lft forever
[root@node2 ~]# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN qlen 1
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: enp0s3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 08:00:27:90:fe:5a brd ff:ff:ff:ff:ff:ff
inet 10.0.2.15/24 brd 10.0.2.255 scope global dynamic enp0s3
valid_lft 86306sec preferred_lft 86306sec
inet6 fe80::4c50:7b9a:e166:b0a9/64 scope link
valid_lft forever preferred_lft forever
3: enp0s8: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 08:00:27:39:0b:1d brd ff:ff:ff:ff:ff:ff
inet6 fe80::1ccc:8e24:f0b6:8de2/64 scope link
valid_lft forever preferred_lft forever
4: enp0s9: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 08:00:27:66:86:7f brd ff:ff:ff:ff:ff:ff
inet6 fe80::bb82:6fa3:a503:b275/64 scope link
valid_lft forever preferred_lft forever
Ethernet Channel Bonding
Le Channel Bonding est un regroupement d'interfaces réseau sur le même serveur afin de mettre en place la redondance ou d'augmenter les performances.
Le Channel Bonding est géré nativement sous Linux. Aucune application tierce n'est requise.
Configuration du node1.opensvc.loc
Assurez-vous que le module bonding soit chargé :
[root@node1 ~]# lsmod | grep bonding [root@node1 ~]# modprobe bonding [root@node1 ~]# lsmod | grep bonding bonding 145728 0
Consultez la configuration des interfaces réseaux :
[root@node1 ~]# nmcli c show NAME UUID TYPE DEVICE Wired connection 1 b24208cf-9b3d-3b5f-b9ea-95668778104f 802-3-ethernet enp0s3 Wired connection 2 d3aa4096-a44f-39e2-adb8-a59c542ed02b 802-3-ethernet -- Wired connection 3 d9159b5d-0d16-3930-9662-b0c1b90f2450 802-3-ethernet --
Créez le fichier /etc/sysconfig/network-scripts/ifcfg-bond0 :
[root@node1 ~]# vi /etc/sysconfig/network-scripts/ifcfg-bond0 [root@node1 ~]# cat /etc/sysconfig/network-scripts/ifcfg-bond0 DEVICE=bond0 USERCTL=no BOOTPROTO=none ONBOOT=yes IPADDR=192.168.121.1 NETMASK=255.255.255.0 NETWORK=192.168.121.0 BONDING_OPTS="miimon=100 mode=balance-xor" TYPE=Unknown IPV6INIT=no
Créez le fichier /etc/sysconfig/network-scripts/ifcfg-enp0s8 :
[root@node1 ~]# vi /etc/sysconfig/network-scripts/ifcfg-enp0s8 [root@node1 ~]# cat /etc/sysconfig/network-scripts/ifcfg-enp0s8 DEVICE=enp0s8 BOOTPROTO=none ONBOOT=yes MASTER=bond0 SLAVE=yes USERCTL=no
Créez le fichier /etc/sysconfig/network-scripts/ifcfg-enp0s9 :
[root@node1 ~]# vi /etc/sysconfig/network-scripts/ifcfg-enp0s9 [root@node1 ~]# cat /etc/sysconfig/network-scripts/ifcfg-enp0s9 DEVICE=enp0s9 BOOTPROTO=none ONBOOT=yes MASTER=bond0 SLAVE=yes USERCTL=no
Créez le fichier /etc/modprobe.d/bonding.conf :
[root@node1 ~]# vi /etc/modprobe.d/bonding.conf [root@node1 ~]# cat /etc/modprobe.d/bonding.conf alias bond0 bonding
Modifiez le fichier /etc/hosts :
[root@node1 ~]# vi /etc/hosts [root@node1 ~]# cat /etc/hosts 127.0.0.1 localhost.localdomain localhost ::1 localhost6.localdomain6 localhost6 192.168.121.1 node1.opensvc.loc node1 192.168.121.2 node2.opensvc.loc node2 192.168.121.42 svc1.opensvc.loc svc1
Re-démarrez le service network :
[root@node1 ~]# nmcli c show NAME UUID TYPE DEVICE System enp0s8 00cb8299-feb9-55b6-a378-3fdc720e0bc6 802-3-ethernet enp0s8 System enp0s9 93d13955-e9e2-a6bd-df73-12e3c747f122 802-3-ethernet enp0s9 Wired connection 1 4ea414aa-a08a-3a0d-8d8c-09aca9bd7fcc 802-3-ethernet enp0s3 bond0 4e7fb7b3-d782-4952-b7dc-bc26cf1f5e9a bond bond0 Wired connection 2 d6c9febc-845b-3e4b-91eb-a9f001670cdc 802-3-ethernet -- Wired connection 3 b9978cef-7b73-318f-aeeb-454d9a1ffeeb 802-3-ethernet --
Vérifiez la configuration du réseau :
[root@node1 ~]# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN qlen 1
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: enp0s3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 08:00:27:41:5b:d7 brd ff:ff:ff:ff:ff:ff
inet 10.0.2.15/24 brd 10.0.2.255 scope global dynamic enp0s3
valid_lft 85360sec preferred_lft 85360sec
inet6 fe80::647b:5dad:397d:3ebb/64 scope link
valid_lft forever preferred_lft forever
3: enp0s8: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master bond0 state UP qlen 1000
link/ether 08:00:27:59:5a:01 brd ff:ff:ff:ff:ff:ff
4: enp0s9: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master bond0 state UP qlen 1000
link/ether 08:00:27:59:5a:01 brd ff:ff:ff:ff:ff:ff
5: bond0: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP qlen 1000
link/ether 08:00:27:59:5a:01 brd ff:ff:ff:ff:ff:ff
inet 192.168.121.1/24 brd 192.168.121.255 scope global bond0
valid_lft forever preferred_lft forever
inet6 fe80::a00:27ff:fe59:5a01/64 scope link tentative dadfailed
valid_lft forever preferred_lft forever
Consultez la configuration de l'interface bond0 :
[root@node1 ~]# cat /proc/net/bonding/bond0 Ethernet Channel Bonding Driver: v3.7.1 (April 27, 2011) Bonding Mode: load balancing (xor) Transmit Hash Policy: layer2 (0) MII Status: up MII Polling Interval (ms): 100 Up Delay (ms): 0 Down Delay (ms): 0 Slave Interface: enp0s8 MII Status: up Speed: 1000 Mbps Duplex: full Link Failure Count: 0 Permanent HW addr: 08:00:27:59:5a:01 Slave queue ID: 0 Slave Interface: enp0s9 MII Status: up Speed: 1000 Mbps Duplex: full Link Failure Count: 0 Permanent HW addr: 08:00:27:43:66:3e Slave queue ID: 0
Configuration du node2.opensvc.loc
Assurez-vous que le module bonding soit chargé :
[root@node2 ~]# lsmod | grep bonding [root@node2 ~]# modprobe bonding [root@node2 ~]# lsmod | grep bonding bonding 145728 0
Consultez la configuration des interfaces réseaux :
[root@node2 ~]# nmcli c show NAME UUID TYPE DEVICE Wired connection 1 794e172c-62ad-3e46-9ee4-ac8f28a250e3 802-3-ethernet enp0s3 Wired connection 2 6f91b884-972c-340c-a61c-7a2a67305ad9 802-3-ethernet enp0s8 Wired connection 3 4b7cbfcd-0558-3000-983b-39db9485f7ae 802-3-ethernet enp0s9
Créez le fichier /etc/sysconfig/network-scripts/ifcfg-bond0 :
[root@node2 ~]# vi /etc/sysconfig/network-scripts/ifcfg-bond0 [root@node2 ~]# cat /etc/sysconfig/network-scripts/ifcfg-bond0 DEVICE=bond0 USERCTL=no BOOTPROTO=none ONBOOT=yes IPADDR=192.168.121.2 NETMASK=255.255.255.0 NETWORK=192.168.121.0 BONDING_OPTS="miimon=100 mode=balance-xor" TYPE=Unknown IPV6INIT=no
Créez le fichier /etc/sysconfig/network-scripts/ifcfg-enp0s8 :
[root@node2 ~]# vi /etc/sysconfig/network-scripts/ifcfg-enp0s8 [root@node2 ~]# cat /etc/sysconfig/network-scripts/ifcfg-enp0s8 DEVICE=enp0s8 BOOTPROTO=none ONBOOT=yes MASTER=bond0 SLAVE=yes USERCTL=no
Créez le fichier /etc/sysconfig/network-scripts/ifcfg-enp0s9 :
[root@node2 ~]# vi /etc/sysconfig/network-scripts/ifcfg-enp0s9 [root@node2 ~]# cat /etc/sysconfig/network-scripts/ifcfg-enp0s9 DEVICE=enp0s9 BOOTPROTO=none ONBOOT=yes MASTER=bond0 SLAVE=yes USERCTL=no
Créez le fichier /etc/modprobe.d/bonding.conf :
[root@node2 ~]# vi /etc/modprobe.d/bonding.conf [root@node2 ~]# cat /etc/modprobe.d/bonding.conf alias bond0 bonding
Modifiez le fichier /etc/hosts :
[root@node2 ~]# vi /etc/hosts [root@node2 ~]# cat /etc/hosts 127.0.0.1 localhost.localdomain localhost ::1 localhost6.localdomain6 localhost6 192.168.121.1 node1.opensvc.loc node1 192.168.121.2 node2.opensvc.loc node2 192.168.121.42 svc1.opensvc.loc svc1
Re-démarrez le service network :
[root@node2 ~]# systemctl restart network [root@node2 ~]# nmcli c show NAME UUID TYPE DEVICE System enp0s8 00cb8299-feb9-55b6-a378-3fdc720e0bc6 802-3-ethernet enp0s8 System enp0s9 93d13955-e9e2-a6bd-df73-12e3c747f122 802-3-ethernet enp0s9 Wired connection 1 794e172c-62ad-3e46-9ee4-ac8f28a250e3 802-3-ethernet enp0s3 bond0 8ee04199-caad-493f-80b3-003b16bf640e bond bond0 Wired connection 2 6f91b884-972c-340c-a61c-7a2a67305ad9 802-3-ethernet -- Wired connection 3 4b7cbfcd-0558-3000-983b-39db9485f7ae 802-3-ethernet --
Vérifiez la configuration du réseau :
[root@node2 ~]# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN qlen 1
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: enp0s3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 08:00:27:13:13:33 brd ff:ff:ff:ff:ff:ff
inet 10.0.2.15/24 brd 10.0.2.255 scope global dynamic enp0s3
valid_lft 84965sec preferred_lft 84965sec
inet6 fe80::2761:4ee8:d2de:d0e2/64 scope link
valid_lft forever preferred_lft forever
3: enp0s8: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master bond0 state UP qlen 1000
link/ether 08:00:27:47:75:77 brd ff:ff:ff:ff:ff:ff
4: enp0s9: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master bond0 state UP qlen 1000
link/ether 08:00:27:47:75:77 brd ff:ff:ff:ff:ff:ff
5: bond0: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP qlen 1000
link/ether 08:00:27:47:75:77 brd ff:ff:ff:ff:ff:ff
inet 192.168.121.2/24 brd 192.168.121.255 scope global bond0
valid_lft forever preferred_lft forever
inet6 fe80::a00:27ff:fe47:7577/64 scope link
valid_lft forever preferred_lft forever
Consultez la configuration de l'interface bond0 :
[root@node2 ~]# cat /proc/net/bonding/bond0 Ethernet Channel Bonding Driver: v3.7.1 (April 27, 2011) Bonding Mode: load balancing (xor) Transmit Hash Policy: layer2 (0) MII Status: up MII Polling Interval (ms): 100 Up Delay (ms): 0 Down Delay (ms): 0 Slave Interface: enp0s9 MII Status: up Speed: 1000 Mbps Duplex: full Link Failure Count: 0 Permanent HW addr: 08:00:27:47:75:77 Slave queue ID: 0 Slave Interface: enp0s8 MII Status: up Speed: 1000 Mbps Duplex: full Link Failure Count: 0 Permanent HW addr: 08:00:27:5a:91:89 Slave queue ID: 0
Mise en place des LVM de test
Configuration du node1.opensvc.loc ET node2.opensvc.loc
Créez un fichier spécial de type bloc (b) ayant un majeur de 7 et un mineur de 0 et les permissions (-m) de 660 :
[root@node1 ~]# mknod -m 0660 /dev/loop0 b 7 0
[root@node2 ~]# mknod -m 0660 /dev/loop0 b 7 0
Associez le périphérique /dev/loop0 avec le fichier spécial /dev/sda1 :
[root@node1 ~]# losetup /dev/loop0 /dev/sda1
[root@node2 ~]# losetup /dev/loop0 /dev/sda1
Initialisez le LVM avec la commande vgscan :
[root@node1 ~]# vgscan Reading volume groups from cache.
[root@node2 ~]# vgscan Reading volume groups from cache.
Créez ensuite un PV sur /dev/loop0 :
[root@node1 ~]# pvcreate /dev/loop0 WARNING: xfs signature detected on /dev/loop0 at offset 0. Wipe it? [y/n]: n Aborted wiping of xfs. 1 existing signature left on the device. [root@node1 ~]# [root@node1 ~]# pvcreate /dev/loop0 WARNING: xfs signature detected on /dev/loop0 at offset 0. Wipe it? [y/n]: y Wiping xfs signature on /dev/loop0. Physical volume "/dev/loop0" successfully created.
[root@node2 ~]# pvcreate /dev/loop0 WARNING: xfs signature detected on /dev/loop0 at offset 0. Wipe it? [y/n]: y Wiping xfs signature on /dev/loop0. Physical volume "/dev/loop0" successfully created.
Créez un LV dénommé vgsvc1 sur /dev/loop0 :
[root@node1 ~]# vgcreate -s 4M vgsvc1 /dev/loop0 Volume group "vgsvc1" successfully created [root@node1 ~]# [root@node1 ~]# vgdisplay --- Volume group --- VG Name vgsvc1 System ID Format lvm2 Metadata Areas 1 Metadata Sequence No 1 VG Access read/write VG Status resizable MAX LV 0 Cur LV 0 Open LV 0 Max PV 0 Cur PV 1 Act PV 1 VG Size 196.00 MiB PE Size 4.00 MiB Total PE 49 Alloc PE / Size 0 / 0 Free PE / Size 49 / 196.00 MiB VG UUID hhvl4f-avgL-2a5z-iT4G-9fMm-KpdC-xXz31C
[root@node2 ~]# vgcreate -s 4M vgsvc1 /dev/loop0 Volume group "vgsvc1" successfully created [root@node2 ~]# [root@node2 ~]# vgdisplay --- Volume group --- VG Name vgsvc1 System ID Format lvm2 Metadata Areas 1 Metadata Sequence No 1 VG Access read/write VG Status resizable MAX LV 0 Cur LV 0 Open LV 0 Max PV 0 Cur PV 1 Act PV 1 VG Size 196.00 MiB PE Size 4.00 MiB Total PE 49 Alloc PE / Size 0 / 0 Free PE / Size 49 / 196.00 MiB VG UUID hhvl4f-avgL-2a5z-iT4G-9fMm-KpdC-xXz31C
Créez deux LV de 96 Mo dénommés lvappsvc1 et lvdatasvc1 respectivement :
[root@node1 ~]# lvcreate -L 95 -n lvappsvc1 vgsvc1 Rounding up size to full physical extent 96.00 MiB Logical volume "lvappsvc1" created. [root@node1 ~]# [root@node1 ~]# lvcreate -L 95 -n lvdatasvc1 vgsvc1 Rounding up size to full physical extent 96.00 MiB Logical volume "lvdatasvc1" created. [root@node1 ~]# [root@node1 ~]# lvdisplay --- Logical volume --- LV Path /dev/vgsvc1/lvappsvc1 LV Name lvappsvc1 VG Name vgsvc1 LV UUID LiIKMj-eE9N-dfxu-rjv1-F98Z-8Pr0-xCDdO5 LV Write Access read/write LV Creation host, time node1.opensvc.loc, 2018-09-12 07:28:47 +0200 LV Status available # open 0 LV Size 96.00 MiB Current LE 24 Segments 1 Allocation inherit Read ahead sectors auto - currently set to 8192 Block device 253:0 --- Logical volume --- LV Path /dev/vgsvc1/lvdatasvc1 LV Name lvdatasvc1 VG Name vgsvc1 LV UUID pisl34-3MJP-0ZxT-y7iO-FKG9-qbxG-7doXsS LV Write Access read/write LV Creation host, time node1.opensvc.loc, 2018-09-12 07:29:54 +0200 LV Status available # open 0 LV Size 96.00 MiB Current LE 24 Segments 1 Allocation inherit Read ahead sectors auto - currently set to 8192 Block device 253:1
[root@node2 ~]# lvcreate -L 95 -n lvappsvc1 vgsvc1 Rounding up size to full physical extent 96.00 MiB Logical volume "lvappsvc1" created. [root@node2 ~]# [root@node2 ~]# lvcreate -L 95 -n lvdatasvc1 vgsvc1 Rounding up size to full physical extent 96.00 MiB Logical volume "lvdatasvc1" created. [root@node2 ~]# [root@node2 ~]# lvdisplay --- Logical volume --- LV Path /dev/vgsvc1/lvappsvc1 LV Name lvappsvc1 VG Name vgsvc1 LV UUID LiIKMj-eE9N-dfxu-rjv1-F98Z-8Pr0-xCDdO5 LV Write Access read/write LV Creation host, time node2.opensvc.loc, 2018-09-12 07:28:47 +0200 LV Status available # open 0 LV Size 96.00 MiB Current LE 24 Segments 1 Allocation inherit Read ahead sectors auto - currently set to 8192 Block device 253:0 --- Logical volume --- LV Path /dev/vgsvc1/lvdatasvc1 LV Name lvdatasvc1 VG Name vgsvc1 LV UUID pisl34-3MJP-0ZxT-y7iO-FKG9-qbxG-7doXsS LV Write Access read/write LV Creation host, time node2.opensvc.loc, 2018-09-12 07:29:54 +0200 LV Status available # open 0 LV Size 96.00 MiB Current LE 24 Segments 1 Allocation inherit Read ahead sectors auto - currently set to 8192 Block device 253:1
Créez un système de fichiers ext4 sur chaque LV :
[root@node1 ~]# mkfs.ext4 /dev/vgsvc1/lvappsvc1 mke2fs 1.42.9 (28-Dec-2013) Filesystem label= OS type: Linux Block size=1024 (log=0) Fragment size=1024 (log=0) Stride=0 blocks, Stripe width=0 blocks 24576 inodes, 98304 blocks 4915 blocks (5.00%) reserved for the super user First data block=1 Maximum filesystem blocks=33685504 12 block groups 8192 blocks per group, 8192 fragments per group 2048 inodes per group Superblock backups stored on blocks: 8193, 24577, 40961, 57345, 73729 Allocating group tables: done Writing inode tables: done Creating journal (4096 blocks): done Writing superblocks and filesystem accounting information: done [root@node1 ~]# [root@node1 ~]# mkfs.ext4 /dev/vgsvc1/lvdatasvc1 mke2fs 1.42.9 (28-Dec-2013) Filesystem label= OS type: Linux Block size=1024 (log=0) Fragment size=1024 (log=0) Stride=0 blocks, Stripe width=0 blocks 24576 inodes, 98304 blocks 4915 blocks (5.00%) reserved for the super user First data block=1 Maximum filesystem blocks=33685504 12 block groups 8192 blocks per group, 8192 fragments per group 2048 inodes per group Superblock backups stored on blocks: 8193, 24577, 40961, 57345, 73729 Allocating group tables: done Writing inode tables: done Creating journal (4096 blocks): done Writing superblocks and filesystem accounting information: done
[root@node2 ~]# mkfs.ext4 /dev/vgsvc1/lvappsvc1 mke2fs 1.42.9 (28-Dec-2013) Filesystem label= OS type: Linux Block size=1024 (log=0) Fragment size=1024 (log=0) Stride=0 blocks, Stripe width=0 blocks 24576 inodes, 98304 blocks 4915 blocks (5.00%) reserved for the super user First data block=1 Maximum filesystem blocks=33685504 12 block groups 8192 blocks per group, 8192 fragments per group 2048 inodes per group Superblock backups stored on blocks: 8193, 24577, 40961, 57345, 73729 Allocating group tables: done Writing inode tables: done Creating journal (4096 blocks): done Writing superblocks and filesystem accounting information: done [root@node2 ~]# [root@node2 ~]# mkfs.ext4 /dev/vgsvc1/lvdatasvc1 mke2fs 1.42.9 (28-Dec-2013) Filesystem label= OS type: Linux Block size=1024 (log=0) Fragment size=1024 (log=0) Stride=0 blocks, Stripe width=0 blocks 24576 inodes, 98304 blocks 4915 blocks (5.00%) reserved for the super user First data block=1 Maximum filesystem blocks=33685504 12 block groups 8192 blocks per group, 8192 fragments per group 2048 inodes per group Superblock backups stored on blocks: 8193, 24577, 40961, 57345, 73729 Allocating group tables: done Writing inode tables: done Creating journal (4096 blocks): done Writing superblocks and filesystem accounting information: done
Créez les points de montage /svc1/data et /svc1/app :
[root@node1 ~]# mkdir -p /svc1/data [root@node1 ~]# mkdir /svc1/app
[root@node2 ~]# mkdir -p /svc1/data [root@node2 ~]# mkdir /svc1/app
Dernièrement, montez les deux systèmes de fichiers :
[root@node1 ~]# mount -t ext4 /dev/vgsvc1/lvappsvc1 /svc1/app [root@node1 ~]# [root@node1 ~]# mount -t ext4 /dev/vgsvc1/lvdatasvc1 /svc1/data
[root@node2 ~]# mount -t ext4 /dev/vgsvc1/lvappsvc1 /svc1/app [root@node2 ~]# [root@node2 ~]# mount -t ext4 /dev/vgsvc1/lvdatasvc1 /svc1/data
Installation d'opensvc-agent
Installation sur node1.opensvc.loc et node2.opensvc.loc
Téléchargez la dernière version d'opensvc-agent :
[root@node1 ~]# wget -O /tmp/opensvc.latest.rpm https://repo.opensvc.com/rpms/current --2018-09-12 05:01:14-- https://repo.opensvc.com/rpms/current Resolving repo.opensvc.com (repo.opensvc.com)... 37.59.71.9 Connecting to repo.opensvc.com (repo.opensvc.com)|37.59.71.9|:443... connected. HTTP request sent, awaiting response... 200 OK Length: 1248463 (1.2M) Saving to: ‘/tmp/opensvc.latest.rpm’ 100%[======================================>] 1,248,463 2.86MB/s in 0.4s 2018-09-12 05:01:15 (2.86 MB/s) - ‘/tmp/opensvc.latest.rpm’ saved [1248463/1248463]
[root@node2 ~]# wget -O /tmp/opensvc.latest.rpm https://repo.opensvc.com/rpms/current --2018-09-12 05:02:13-- https://repo.opensvc.com/rpms/current Resolving repo.opensvc.com (repo.opensvc.com)... 37.59.71.9 Connecting to repo.opensvc.com (repo.opensvc.com)|37.59.71.9|:443... connected. HTTP request sent, awaiting response... 200 OK Length: 1248463 (1.2M) Saving to: ‘/tmp/opensvc.latest.rpm’ 100%[======================================>] 1,248,463 296KB/s in 4.1s 2018-09-12 05:02:17 (296 KB/s) - ‘/tmp/opensvc.latest.rpm’ saved [1248463/1248463]
Installez l'opensvc-agent :
[root@node1 ~]# yum -y install /tmp/opensvc.latest.rpm
[root@node2 ~]# yum -y install /tmp/opensvc.latest.rpm
Démarrez et activez le service opensvc-agent :
[root@node1 ~]# systemctl start opensvc-agent.service [root@node1 ~]# systemctl is-active opensvc-agent.service active
[root@node2 ~]# systemctl start opensvc-agent.service [root@node2 ~]# systemctl is-active opensvc-agent.service active
Vérifiez le statut du service opensvc-agent :
[root@node1 ~]# systemctl status opensvc-agent
● opensvc-agent.service - OpenSVC agent
Loaded: loaded (/usr/lib/systemd/system/opensvc-agent.service; enabled; vendor preset: disabled)
Active: active (running) since Wed 2018-09-12 05:03:30 CEST; 5min ago
Docs: http://docs.opensvc.com/
file:/usr/share/doc/opensvc/
man:nodemgr(1)
man:svcmgr(1)
man:svcmon(1)
Process: 7532 ExecStart=/usr/share/opensvc/bin/nodemgr daemon start (code=exited, status=0/SUCCESS)
Main PID: 7555 (python)
CGroup: /system.slice/opensvc-agent.service
└─7555 /usr/bin/python /usr/share/opensvc/lib/osvcd.py
Sep 12 05:03:30 node1.opensvc.loc python[7555]: node1.opensvc.loc.osvcd.moni...e
Sep 12 05:03:30 node1.opensvc.loc python[7555]: node1.opensvc.loc.osvcd.moni...h
Sep 12 05:03:31 node1.opensvc.loc python[7555]: node1.opensvc.loc.osvcd.moni...n
Sep 12 05:03:31 node1.opensvc.loc python[7555]: node1.opensvc.loc.osvcd.moni...r
Sep 12 05:03:31 node1.opensvc.loc python[7555]: node1.opensvc.loc.osvcd.moni...e
Sep 12 05:03:31 node1.opensvc.loc python[7555]: node1.opensvc.loc.osvcd.moni...e
Sep 12 05:03:32 node1.opensvc.loc python[7555]: node1.opensvc.loc.osvcd node...)
Sep 12 05:03:32 node1.opensvc.loc python[7555]: node1.opensvc.loc.osvcd.moni...e
Sep 12 05:03:33 node1.opensvc.loc python[7555]: node1.opensvc.loc.osvcd.list...e
Sep 12 05:04:34 node1.opensvc.loc python[7555]: node1.opensvc.loc.osvcd.sche...5
Hint: Some lines were ellipsized, use -l to show in full.
[root@node2 ~]# systemctl status opensvc-agent
● opensvc-agent.service - OpenSVC agent
Loaded: loaded (/usr/lib/systemd/system/opensvc-agent.service; enabled; vendor preset: disabled)
Active: active (running) since Wed 2018-09-12 05:03:53 CEST; 3min 29s ago
Docs: http://docs.opensvc.com/
file:/usr/share/doc/opensvc/
man:nodemgr(1)
man:svcmgr(1)
man:svcmon(1)
Process: 7655 ExecStart=/usr/share/opensvc/bin/nodemgr daemon start (code=exited, status=0/SUCCESS)
Main PID: 7687 (python)
CGroup: /system.slice/opensvc-agent.service
└─7687 /usr/bin/python /usr/share/opensvc/lib/osvcd.py
Sep 12 05:03:53 node2.opensvc.loc python[7687]: node2.opensvc.loc.osvcd.moni...h
Sep 12 05:03:54 node2.opensvc.loc python[7687]: node2.opensvc.loc.osvcd.moni...n
Sep 12 05:03:54 node2.opensvc.loc python[7687]: node2.opensvc.loc.osvcd.moni...r
Sep 12 05:03:54 node2.opensvc.loc python[7687]: node2.opensvc.loc.osvcd.moni...e
Sep 12 05:03:54 node2.opensvc.loc python[7687]: node2.opensvc.loc.osvcd.moni...e
Sep 12 05:03:55 node2.opensvc.loc python[7687]: node2.opensvc.loc.osvcd node...)
Sep 12 05:03:55 node2.opensvc.loc python[7687]: node2.opensvc.loc.osvcd.moni...e
Sep 12 05:03:56 node2.opensvc.loc python[7687]: node2.opensvc.loc.osvcd.list...e
Sep 12 05:04:56 node2.opensvc.loc python[7687]: node2.opensvc.loc.osvcd.sche...3
Sep 12 05:06:46 node2.opensvc.loc python[7687]: node2.opensvc.loc.osvcd.sche...8
Hint: Some lines were ellipsized, use -l to show in full.
Utilisez la commande svcmon pour visualiser le statut des services sur le noeud :
[root@node1 ~]# svcmon Threads node1.opensvc.loc listener running 0.0.0.0:1214 monitor running scheduler running 6 wait Nodes node1.opensvc.loc score | 92 load 15m | 0.1 mem | 50/98%:488m swap | 1/90%:1g state | Services node1.opensvc.loc
[root@node2 ~]# svcmon Threads node2.opensvc.loc listener running 0.0.0.0:1214 monitor running scheduler running 6 wait Nodes node2.opensvc.loc score | 92 load 15m | 0.1 mem | 50/98%:488m swap | 1/90%:1g state | Services node2.opensvc.loc
L'agent utilise des threads pour gérer des rôles différents :
- Listener - à l'écoute de requêtes des daemon d'autres neouds, du collector, de nodemgr et de svcmgr,
- Monitor - gère les actions globales de démarrage du cluster et de failover,
- Scheduler - gère la gestion des tâches sur les neouds,
- Heartbeat - vérifie si les autres neouds sont “en vie” et échange des informations entre les neouds.
Pour mettre en place un heartbeat, saisissez la commande suivante :
[root@node1 ~]# nodemgr set --param hb#1.type --value unicast [root@node1 ~]# svcmon Threads node1.opensvc.loc hb#1.rx running 0.0.0.0:10000 | / hb#1.tx running | / listener running 0.0.0.0:1214 monitor running scheduler running 5 wait Nodes node1.opensvc.loc score | 92 load 15m | 0.1 mem | 51/98%:488m swap | 1/90%:1g state | Services node1.opensvc.loc
La configuration de chaque noeud est stockée dans le fichier /etc/opensvc/node.conf.
[root@node1 ~]# cat /etc/opensvc/node.conf [cluster] id = 6de28056-b638-11e8-bb2d-080027415bd7 nodes = node1.opensvc.loc name = default secret = 6e828240b63811e8819b080027415bd7 [hb#1] type = unicast
Les deux nodes sont en mode TST par défaut. En production, le mode des noeuds doit être modifié à PRD de la façon suivante :
# nodemgr set --param node.env --value PRD
Clefs SSH
Les membres du cluster ont besoin d'effectuer des authentifications mutuelles par ssh afin d'échanger des fichiers de configuration. Chaque neoud utilisera une authentification par clefs :
- node1 se connectera à node2 en tant que root,
- node2 se connectera à node1 en tant que root.
Transférez donc la clef de node1 à node2 :
[root@node1 ~]# ssh-copy-id root@node2 /bin/ssh-copy-id: INFO: Source of key(s) to be installed: "/root/.ssh/id_rsa.pub" The authenticity of host 'node2 (192.168.121.2)' can't be established. ECDSA key fingerprint is SHA256:RgOsp/XI7JHNq+oIfHKw+jkHdtTnBIh+Dd7kVmHRxtU. ECDSA key fingerprint is MD5:19:cd:05:58:af:2c:10:82:52:ba:e3:31:df:bd:72:54. Are you sure you want to continue connecting (yes/no)? yes /bin/ssh-copy-id: INFO: attempting to log in with the new key(s), to filter out any that are already installed /bin/ssh-copy-id: INFO: 1 key(s) remain to be installed -- if you are prompted now it is to install the new keys root@node2's password: fenestros Number of key(s) added: 1 Now try logging into the machine, with: "ssh 'root@node2'" and check to make sure that only the key(s) you wanted were added. [root@node1 ~]# ssh node2 hostname node2.opensvc.loc
Transférez donc la clef de node2 à node1 :
[root@node2 ~]# ssh-copy-id root@node1 /bin/ssh-copy-id: INFO: Source of key(s) to be installed: "/root/.ssh/id_rsa.pub" The authenticity of host 'node1 (192.168.121.1)' can't be established. ECDSA key fingerprint is SHA256:RgOsp/XI7JHNq+oIfHKw+jkHdtTnBIh+Dd7kVmHRxtU. ECDSA key fingerprint is MD5:19:cd:05:58:af:2c:10:82:52:ba:e3:31:df:bd:72:54. Are you sure you want to continue connecting (yes/no)? yes /bin/ssh-copy-id: INFO: attempting to log in with the new key(s), to filter out any that are already installed /bin/ssh-copy-id: INFO: 1 key(s) remain to be installed -- if you are prompted now it is to install the new keys root@node1's password: fenestros Number of key(s) added: 1 Now try logging into the machine, with: "ssh 'root@node1'" and check to make sure that only the key(s) you wanted were added. [root@node2 ~]# ssh node1 hostname node1.opensvc.loc
Création d'un Service
Créez un service dénommé svc1 :
[root@node1 ~]# svcmgr -s svc1 create
Cette commande génère un fichier dans /etc/opensvc appelé svc1.conf :
[root@node1 ~]# cat /etc/opensvc/svc1.conf [DEFAULT] id = 7adc8d25-1668-4078-a3f0-5b4472c436d2
Ce fichier est vide. Il convient donc de l'éditer afin de définir le service.
Créer le fichier de service
Editez donc ce fichier afin de créer un service appelé MyApp qui s'exécute sur node1 avec un failover vers node2. Ce service utilisera une adresse IP dénommé svc1.opensvc.loc, un VG dénommé vgsvc1 et deux systèmes de fichiers sur les LV /dev/mapper/vgsvc1-lvappsvc1 et /dev/mapper/vgsvc1-lvdatasvc1 :
[root@node1 ~]# svcmgr -s svc1 edit config [root@node1 ~]# cat /etc/opensvc/svc1.conf [DEFAULT] id = 7adc8d25-1668-4078-a3f0-5b4472c436d2 app = MyApp nodes = node1.opensvc.loc node2.opensvc.loc [ip#0] ipname = svc1.opensvc.loc ipdev = bond0 [disk#0] type = vg name = vgsvc1 pvs = /dev/loop0 [fs#app] type = ext4 dev = /dev/mapper/vgsvc1-lvappsvc1 mnt = /svc1/app [fs#data] type = ext4 dev = /dev/mapper/vgsvc1-lvdatasvc1 mnt = /svc1/data
Le Répertoire des Scripts de Démarrage des Services
Les services sont utilisés pour gérer des applications. Si, par exemple, on souhaitait construire un service de type LAMP, nous aurons besoins de 2 scripts, un pour la base de données MySQL et l'autre pour le serveur Apache. Ces scripts doivent se trouver dans le répertoire /etc/opensvc/svc1.dir. Cependant, OpenSVC cherche dans le répertoire /etc/opensvc/svc1.d pour des scripts :
[root@node1 ~]# mkdir /etc/opensvc/svc1.dir [root@node1 ~]# ln -s /etc/opensvc/svc1.dir /etc/opensvc/svc1.d
Service Management Facility
Afin de rendre la gestion des services plus facile, OpenSVC crée un lien symbolique svc1 dans le répertoire /etc/opensvc qui pointe vers la commande /usr/bin/svcmgr :
[root@node1 ~]# ls -l /etc/opensvc total 8 -rw-r-----. 1 root root 158 Sep 12 05:20 node.conf lrwxrwxrwx. 1 root root 15 Sep 12 05:23 svc1 -> /usr/bin/svcmgr -rw-r--r--. 1 root root 332 Sep 12 05:39 svc1.conf lrwxrwxrwx. 1 root root 21 Sep 12 06:12 svc1.d -> /etc/opensvc/svc1.dir drwxr-xr-x. 2 root root 6 Sep 12 06:09 svc1.dir
Le service peut donc être maintenant géré de deux façons :
[root@node1 ~]# svcmgr -s svc1 print status
svc1 down warn
`- instances
|- node2.opensvc.loc undef daemon down
`- node1.opensvc.loc down warn,
| frozen,
| idle
|- ip#0 ....... down svc1.opensvc.loc@bond0
|- disk#0 ....... down vg vgsvc1
|- fs#app ....... down ext4 /dev/mapper/vgsvc1-lvapps
| vc1@/svc1/app
|- fs#data ....... down ext4 /dev/mapper/vgsvc1-lvdata
| svc1@/svc1/data
`- sync#i0 ...O./. warn rsync svc config to nodes
warn: passive node needs
update
ou :
[root@node1 ~]# /etc/opensvc/svc1 print status
svc1 down warn
`- instances
|- node2.opensvc.loc undef daemon down
`- node1.opensvc.loc down warn,
| frozen,
| idle
|- ip#0 ....... down svc1.opensvc.loc@bond0
|- disk#0 ....... down vg vgsvc1
|- fs#app ....... down ext4 /dev/mapper/vgsvc1-lvapps
| vc1@/svc1/app
|- fs#data ....... down ext4 /dev/mapper/vgsvc1-lvdata
| svc1@/svc1/data
`- sync#i0 ...O./. warn rsync svc config to nodes
warn: passive node needs
update
Dans la sortie de la commande ci-dessus, le statut global est warn parce que toutes les ressources sont down.
Démarrez le service :
[root@node1 ~]# svc1 start --local node1.opensvc.loc.svc1.ip#0 checking 192.168.121.42 availability node1.opensvc.loc.svc1.ip#0 ifconfig bond0:1 192.168.121.42 netmask 255.255.255.0 up node1.opensvc.loc.svc1.ip#0 send gratuitous arp to announce 192.168.121.42 is at bond0 node1.opensvc.loc.svc1.disk#0 vgchange --addtag @node1.opensvc.loc vgsvc1 node1.opensvc.loc.svc1.disk#0 output: node1.opensvc.loc.svc1.disk#0 Volume group "vgsvc1" successfully changed node1.opensvc.loc.svc1.disk#0 node1.opensvc.loc.svc1.disk#0 vg vgsvc1 is already up node1.opensvc.loc.svc1.fs#app ext4 /dev/mapper/vgsvc1-lvappsvc1@/svc1/app is already mounted node1.opensvc.loc.svc1.fs#data ext4 /dev/mapper/vgsvc1-lvdatasvc1@/svc1/data is already mounted
Cette séquence :
- vérifie si l'adresse 192.168.121.42 n'est pas déjà utilisée,
- active l'adresse IP,
- active le VG si ceci n'est pas déjà fait,
- monte chaque système de fichiers.
Arrêtez maintenat le service :
[root@node1 ~]# svc1 stop node1.opensvc.loc.svc1 stop action requested
Deémarrez le service :
[root@node1 ~]# svc1 start --local node1.opensvc.loc.svc1.ip#0 checking 192.168.121.42 availability node1.opensvc.loc.svc1.ip#0 ifconfig bond0:1 192.168.121.42 netmask 255.255.255.0 up node1.opensvc.loc.svc1.ip#0 send gratuitous arp to announce 192.168.121.42 is at bond0 node1.opensvc.loc.svc1.disk#0 vgchange --addtag @node1.opensvc.loc vgsvc1 node1.opensvc.loc.svc1.disk#0 output: node1.opensvc.loc.svc1.disk#0 Volume group "vgsvc1" successfully changed node1.opensvc.loc.svc1.disk#0 node1.opensvc.loc.svc1.disk#0 vgchange -a y vgsvc1 node1.opensvc.loc.svc1.disk#0 output: node1.opensvc.loc.svc1.disk#0 2 logical volume(s) in volume group "vgsvc1" now active node1.opensvc.loc.svc1.disk#0 node1.opensvc.loc.svc1.fs#app e2fsck -p /dev/mapper/vgsvc1-lvappsvc1 node1.opensvc.loc.svc1.fs#app output: node1.opensvc.loc.svc1.fs#app /dev/mapper/vgsvc1-lvappsvc1: clean, 12/24576 files, 8757/98304 blocks node1.opensvc.loc.svc1.fs#app node1.opensvc.loc.svc1.fs#app mount -t ext4 /dev/mapper/vgsvc1-lvappsvc1 /svc1/app node1.opensvc.loc.svc1.fs#data e2fsck -p /dev/mapper/vgsvc1-lvdatasvc1 node1.opensvc.loc.svc1.fs#data output: node1.opensvc.loc.svc1.fs#data /dev/mapper/vgsvc1-lvdatasvc1: clean, 12/24576 files, 8757/98304 blocks node1.opensvc.loc.svc1.fs#data node1.opensvc.loc.svc1.fs#data mount -t ext4 /dev/mapper/vgsvc1-lvdatasvc1 /svc1/data
Notez que le processus de démarrage a monté les deux LV :
[root@node1 ~]# mount | grep svc1 /dev/mapper/vgsvc1-lvappsvc1 on /svc1/app type ext4 (rw,relatime,seclabel,data=ordered) /dev/mapper/vgsvc1-lvdatasvc1 on /svc1/data type ext4 (rw,relatime,seclabel,data=ordered)
et activé l'adresse IP :
[root@node1 ~]# ip addr list bond0
5: bond0: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP qlen 1000
link/ether 08:00:27:59:5a:01 brd ff:ff:ff:ff:ff:ff
inet 192.168.121.1/24 brd 192.168.121.255 scope global bond0
valid_lft forever preferred_lft forever
inet 192.168.121.42/24 brd 192.168.121.255 scope global secondary bond0:1
valid_lft forever preferred_lft forever
inet6 fe80::a00:27ff:fe59:5a01/64 scope link tentative dadfailed
valid_lft forever preferred_lft forever
Notre service fonctionne maintenat sur node1 :
[root@node1 ~]# svcmgr -s svc1 print status
svc1 up
`- instances
|- node2.opensvc.loc undef daemon down
`- node1.opensvc.loc up frozen,
| idle,
| started
|- ip#0 ....... up svc1.opensvc.loc@bond0
|- disk#0 ....... up vg vgsvc1
|- fs#app ....... up ext4 /dev/mapper/vgsvc1-lvapps
| vc1@/svc1/app
|- fs#data ....... up ext4 /dev/mapper/vgsvc1-lvdata
| svc1@/svc1/data
`- sync#i0 ...O./. n/a rsync svc config to nodes
info: paused, service not up
Intégrer une Application
Applications launcher directory
Déplacez maintenant le répertoire de lancement des applications sur un LV géré par le cluster :
[root@node1 ~]# cd /etc/opensvc/ [root@node1 opensvc]# ls -lart | grep svc1 lrwxrwxrwx. 1 root root 15 Sep 12 05:23 svc1 -> /usr/bin/svcmgr drwxr-xr-x. 2 root root 6 Sep 12 06:09 svc1.dir lrwxrwxrwx. 1 root root 21 Sep 12 06:12 svc1.d -> /etc/opensvc/svc1.dir -rw-r--r--. 1 root root 358 Sep 12 07:44 svc1.conf [root@node1 opensvc]# rm -f svc1.d [root@node1 opensvc]# rmdir svc1.dir [root@node1 opensvc]# mkdir /svc1/app/init.d [root@node1 opensvc]# ln -s /svc1/app/init.d svc1.d [root@node1 opensvc]# ls -lart | grep svc1 lrwxrwxrwx. 1 root root 15 Sep 12 05:23 svc1 -> /usr/bin/svcmgr -rw-r--r--. 1 root root 358 Sep 12 07:44 svc1.conf lrwxrwxrwx. 1 root root 16 Sep 12 08:44 svc1.d -> /svc1/app/init.d
Obtenir le binaire de l'application
Téléchargez l'application mongoose - un serveur web léger :
[root@node1 opensvc]# cd /svc1/app [root@node1 app]# wget -O /svc1/app/webserver https://storage.googleapis.com/google-code-archive-downloads/v2/code.google.com/mongoose/mongoose-lua-sqlite-ssl-static-x86_64-5.1 --2018-09-12 08:45:53-- https://storage.googleapis.com/google-code-archive-downloads/v2/code.google.com/mongoose/mongoose-lua-sqlite-ssl-static-x86_64-5.1 Resolving storage.googleapis.com (storage.googleapis.com)... 172.217.22.144, 2a00:1450:4007:813::2010 Connecting to storage.googleapis.com (storage.googleapis.com)|172.217.22.144|:443... connected. HTTP request sent, awaiting response... 200 OK Length: 2527016 (2.4M) [application/octet-stream] Saving to: ‘/svc1/app/webserver’ 100%[======================================>] 2,527,016 2.75MB/s in 0.9s 2018-09-12 08:45:55 (2.75 MB/s) - ‘/svc1/app/webserver’ saved [2527016/2527016] [root@node1 app]# ls -l /svc1/app/webserver -rw-r--r--. 1 root root 2527016 Jan 23 2016 /svc1/app/webserver
Rendez exécutable le fichier /svc1/app/webserver :
[root@node1 app]# chmod u+x /svc1/app/webserver
Créez le fichier index.html :
[root@node1 app]# cd /svc1/data/ [root@node1 data]# vi index.html [root@node1 data]# cat index.html <html> <body>It Works !</body> </html>
Créez maintenat un script de gestion de l'application :
[root@node1 data]# cd /svc1/app/init.d
[root@node1 init.d]# vi weblauncher
[root@node1 init.d]# cat weblauncher
#!/bin/bash
SVCROOT=/svc1
APPROOT=${SVCROOT}/app
DAEMON=${APPROOT}/webserver
DAEMON_BASE=$(basename $DAEMON)
DAEMONOPTS="-document_root ${SVCROOT}/data -index_files index.html -listening_port 8080"
function status {
pgrep $DAEMON_BASE >/dev/null 2>&1
}
case $1 in
restart)
killall $DAEMON_BASE
$DAEMON
;;
start)
status && {
echo "already started"
exit 0
}
nohup $DAEMON $DAEMONOPTS >> /dev/null 2>&1 &
;;
stop)
killall $DAEMON_BASE
;;
info)
echo "Name: webserver"
;;
status)
status
exit $?
;;
*)
echo "unsupported action: $1" >&2
exit 1
;;
esac
Rendez ce script exécutable :
[root@node1 init.d]# chmod u+x weblauncher
Testez le script :
[root@node1 init.d]# ./weblauncher status [root@node1 init.d]# echo $? 1 [root@node1 init.d]# ./weblauncher start [root@node1 init.d]# ./weblauncher status [root@node1 init.d]# echo $? 0 [root@node1 init.d]# ./weblauncher stop [root@node1 init.d]# ./weblauncher status [root@node1 init.d]# echo $? 1
Editez maintenant le fichier /etc/opensvc/svc1.conf :
[root@node1 ~]# svc1 edit config [root@node1 ~]# cat /etc/opensvc/svc1.conf [DEFAULT] id = 7adc8d25-1668-4078-a3f0-5b4472c436d2 app = MyApp nodes = node1.opensvc.loc node2.opensvc.loc [ip#0] ipname = svc1.opensvc.loc ipdev = bond0 [disk#0] type = vg name = vgsvc1 pvs = /dev/loop0 [fs#app] type = ext4 dev = /dev/mapper/vgsvc1-lvappsvc1 mnt = /svc1/app [fs#data] type = ext4 dev = /dev/mapper/vgsvc1-lvdatasvc1 mnt = /svc1/data [app#web] script = weblauncher start = 10 check = 10 stop = 90
Cette configuration indique à OpenSVC d'appeler le script weblauncher avec l'argument :
- start quand le service OpenSVC démarre,
- stop quand le service OpenSVC s'arrête,
- status quand le service OpenSVC a besoin de connaître le statut de l'application.
Démarrez maintenat le service svc1 :
[root@node1 ~]# svc1 start --local node1.opensvc.loc.svc1.ip#0 192.168.121.42 is already up on bond0 node1.opensvc.loc.svc1.disk#0 vg vgsvc1 is already up node1.opensvc.loc.svc1.fs#app ext4 /dev/mapper/vgsvc1-lvappsvc1@/svc1/app is already mounted node1.opensvc.loc.svc1.fs#data ext4 /dev/mapper/vgsvc1-lvdatasvc1@/svc1/data is already mounted node1.opensvc.loc.svc1.app#web exec /svc1/app/init.d/weblauncher start as user root node1.opensvc.loc.svc1.app#web start done in 0:00:01.011693 ret 0
En regardant le statut du service, on note que l'application a démarré :
[root@node1 ~]# svc1 print status
svc1 up
`- instances
|- node2.opensvc.loc undef daemon down
`- node1.opensvc.loc up frozen, idle, started
|- ip#0 ....... up svc1.opensvc.loc@bond0
|- disk#0 ....... up vg vgsvc1
|- fs#app ....... up ext4 /dev/mapper/vgsvc1-lvappsvc1@/svc1/app
|- fs#data ....... up ext4 /dev/mapper/vgsvc1-lvdatasvc1@/svc1/data
|- app#web ...../. up forking: weblauncher
`- sync#i0 ...O./. up rsync svc config to nodes
Pour vérifier si c'est vraiment le cas, utilisez les commandes suivantes :
[root@node1 ~]# ps auxww | grep web root 22374 0.0 0.0 2540 328 pts/1 S 09:56 0:00 /svc1/app/webserver -document_root /svc1/data -index_files index.html -listening_port 8080 root 23406 0.0 0.1 112660 972 pts/1 R+ 09:58 0:00 grep --color=auto web [root@node1 ~]# wget -qO - http://svc1.opensvc.loc:8080/ <html> <body>It Works !</body> </html>
Arrêtez maintenant le service svc1 :
[root@node1 ~]# svc1 stop --local node1.opensvc.loc.svc1.app#web exec /svc1/app/init.d/weblauncher stop as user root node1.opensvc.loc.svc1.app#web stop done in 0:00:01.016128 ret 0 node1.opensvc.loc.svc1.fs#data umount /svc1/data node1.opensvc.loc.svc1.fs#app umount /svc1/app node1.opensvc.loc.svc1.disk#0 vgchange --deltag @node1.opensvc.loc vgsvc1 node1.opensvc.loc.svc1.disk#0 output: node1.opensvc.loc.svc1.disk#0 Volume group "vgsvc1" successfully changed node1.opensvc.loc.svc1.disk#0 node1.opensvc.loc.svc1.disk#0 vgchange -a n vgsvc1 node1.opensvc.loc.svc1.disk#0 output: node1.opensvc.loc.svc1.disk#0 0 logical volume(s) in volume group "vgsvc1" now active node1.opensvc.loc.svc1.disk#0 node1.opensvc.loc.svc1.ip#0 ifconfig bond0:1 down node1.opensvc.loc.svc1.ip#0 checking 192.168.121.42 availability
Cette commande :
- arrête l'application,
- démonte les systèmes de fichiers,
- désactive le VG,
- désactive l'adresse IP.
Pour vérifier le statut, utilisez la commande suivante :
[root@node1 ~]# svc1 print status
svc1 down
`- instances
|- node2.opensvc.loc undef daemon down
`- node1.opensvc.loc down frozen, idle
|- ip#0 ....... down svc1.opensvc.loc@bond0
|- disk#0 ....... down vg vgsvc1
|- fs#app ....... down ext4 /dev/mapper/vgsvc1-lvappsvc1@/svc1/app
|- fs#data ....... down ext4 /dev/mapper/vgsvc1-lvdatasvc1@/svc1/data
|- app#web ...../. down forking: weblauncher
`- sync#i0 ...O./. up rsync svc config to nodes
Démarrez de nouveau le service :
[root@node1 ~]# svc1 start --local node1.opensvc.loc.svc1.ip#0 checking 192.168.121.42 availability node1.opensvc.loc.svc1.ip#0 ifconfig bond0:1 192.168.121.42 netmask 255.255.255.0 up node1.opensvc.loc.svc1.ip#0 send gratuitous arp to announce 192.168.121.42 is at bond0 node1.opensvc.loc.svc1.disk#0 vgchange --addtag @node1.opensvc.loc vgsvc1 node1.opensvc.loc.svc1.disk#0 output: node1.opensvc.loc.svc1.disk#0 Volume group "vgsvc1" successfully changed node1.opensvc.loc.svc1.disk#0 node1.opensvc.loc.svc1.disk#0 vgchange -a y vgsvc1 node1.opensvc.loc.svc1.disk#0 output: node1.opensvc.loc.svc1.disk#0 2 logical volume(s) in volume group "vgsvc1" now active node1.opensvc.loc.svc1.disk#0 node1.opensvc.loc.svc1.fs#app e2fsck -p /dev/mapper/vgsvc1-lvappsvc1 node1.opensvc.loc.svc1.fs#app output: node1.opensvc.loc.svc1.fs#app /dev/mapper/vgsvc1-lvappsvc1: clean, 15/24576 files, 11228/98304 blocks node1.opensvc.loc.svc1.fs#app node1.opensvc.loc.svc1.fs#app mount -t ext4 /dev/mapper/vgsvc1-lvappsvc1 /svc1/app node1.opensvc.loc.svc1.fs#data e2fsck -p /dev/mapper/vgsvc1-lvdatasvc1 node1.opensvc.loc.svc1.fs#data output: node1.opensvc.loc.svc1.fs#data /dev/mapper/vgsvc1-lvdatasvc1: clean, 13/24576 files, 8758/98304 blocks node1.opensvc.loc.svc1.fs#data node1.opensvc.loc.svc1.fs#data mount -t ext4 /dev/mapper/vgsvc1-lvdatasvc1 /svc1/data node1.opensvc.loc.svc1.app#web exec /svc1/app/init.d/weblauncher start as user root node1.opensvc.loc.svc1.app#web start done in 0:00:01.009806 ret 0
Service Failover
Le service s'exécute sur node1.
Ajoutez maintenant node2 au cluster de node1 :
[root@node1 ~]# nodemgr get --param cluster.secret 6e828240b63811e8819b080027415bd7 [root@node2 ~]# nodemgr daemon join --secret 6e828240b63811e8819b080027415bd7 --node node1.opensvc.loc node2.opensvc.loc freeze local node node2.opensvc.loc update node.env => TST node2.opensvc.loc update heartbeat hb#1 node2.opensvc.loc join node node1.opensvc.loc node2.opensvc.loc thaw local node
Notez maintentant que les deux noeuds se trouvent dans le cluster :
[root@node2 ~]# svcmon Threads node1.opensvc.loc node2.opensvc.loc hb#1.rx running 0.0.0.0:10000 | O / hb#1.tx running | O / listener running 0.0.0.0:1214 monitor running scheduler running Nodes node1.opensvc.loc node2.opensvc.loc score | 91 92 load 15m | 0.1 0.1 mem | 56/98%:488m 53/98%:488m swap | 1/90%:1g 1/90%:1g state | Services node1.opensvc.loc node2.opensvc.loc svc1 up! failover | O* !!P
Synchronisez les deux noeuds :
[root@node1 ~]# svc1 sync nodes node1.opensvc.loc.svc1.sync#i0 transfered 279 B at 598 B/s node1.opensvc.loc.svc1 request action 'postsync --waitlock=3600' on node node2.opensvc.loc
La réplication n'a lieu que dans le où les conditions suivantes soient remplies :
- le nouveau noeud est déclaré dans le fichier /etc/opensvc/svc1.conf,
- le neoud récepteur a confiance dans le noeud émetteur ( via ssh ),
- le service svc1 soit activé et fonctionnel sur le node1,
- la synchronisation précédente est plus ancienne.
Consultez maintenant le contenu du répertoire /etc/opensvc sur node2 :
[root@node2 ~]# ls -l /etc/opensvc total 8 -rw-------. 1 root root 195 Sep 12 10:21 node.conf lrwxrwxrwx. 1 root root 15 Sep 12 05:23 svc1 -> /usr/bin/svcmgr -rw-r--r--. 1 root root 421 Sep 12 09:52 svc1.conf lrwxrwxrwx. 1 root root 16 Sep 12 08:44 svc1.d -> /svc1/app/init.d drwxr-xr-x. 2 root root 6 Sep 12 06:09 svc1.dir
Notez que lien symbolique orphelin svc1.d clignote.
En comparant ce contenu avec celui du node1, on constate l'absence du répertoire svc1.dir :
[root@node1 ~]# ls -l /etc/opensvc total 8 -rw-r-----. 1 root root 176 Sep 12 10:11 node.conf lrwxrwxrwx. 1 root root 15 Sep 12 05:23 svc1 -> /usr/bin/svcmgr -rw-r--r--. 1 root root 421 Sep 12 09:52 svc1.conf lrwxrwxrwx. 1 root root 16 Sep 12 08:44 svc1.d -> /svc1/app/init.d
Supprimez donc le répertoire svc1.dir sur node2 et créez le répertoire /svc1/app/init.d :
[root@node2 opensvc]# rmdir svc1.dir [root@node2 opensvc]# mkdir /svc1/app/init.d [root@node2 opensvc]# ls -lart | grep svc1 lrwxrwxrwx. 1 root root 15 Sep 12 05:23 svc1 -> /usr/bin/svcmgr lrwxrwxrwx. 1 root root 16 Sep 12 08:44 svc1.d -> /svc1/app/init.d -rw-r--r--. 1 root root 421 Sep 12 09:52 svc1.conf
Installez maintenant l'application mongoose sur node2 :
[root@node2 opensvc]# cd /svc1/app [root@node2 app]# wget -O /svc1/app/webserver https://storage.googleapis.com/google-code-archive-downloads/v2/code.google.com/mongoose/mongoose-lua-sqlite-ssl-static-x86_64-5.1 --2018-09-12 11:36:24-- https://storage.googleapis.com/google-code-archive-downloads/v2/code.google.com/mongoose/mongoose-lua-sqlite-ssl-static-x86_64-5.1 Resolving storage.googleapis.com (storage.googleapis.com)... 216.58.204.240, 2a00:1450:4007:80c::2010 Connecting to storage.googleapis.com (storage.googleapis.com)|216.58.204.240|:443... connected. HTTP request sent, awaiting response... 200 OK Length: 2527016 (2.4M) [application/octet-stream] Saving to: ‘/svc1/app/webserver’ 100%[======================================>] 2,527,016 2.45MB/s in 1.0s 2018-09-12 11:36:25 (2.45 MB/s) - ‘/svc1/app/webserver’ saved [2527016/2527016] [root@node2 app]# ls -l /svc1/app/webserver -rw-r--r--. 1 root root 2527016 Jan 23 2016 /svc1/app/webserver
Rendez /svc1/app/webserver exécutable :
[root@node2 app]# chmod u+x /svc1/app/webserver
Créez le fichier index.html :
[root@node2 app]# cd /svc1/data/ [root@node2 data]# vi index.html [root@node2 data]# cat index.html <html> <body>It Works !</body> </html>
ainsi que le fichier /svc1/app/init.d/weblauncher :
[root@node2 data]# cd /svc1/app/init.d
[root@node2 init.d]# vi weblauncher
[root@node2 init.d]# cat weblauncher
#!/bin/bash
SVCROOT=/svc1
APPROOT=${SVCROOT}/app
DAEMON=${APPROOT}/webserver
DAEMON_BASE=$(basename $DAEMON)
DAEMONOPTS="-document_root ${SVCROOT}/data -index_files index.html -listening_port 8080"
function status {
pgrep $DAEMON_BASE >/dev/null 2>&1
}
case $1 in
restart)
killall $DAEMON_BASE
$DAEMON
;;
start)
status && {
echo "already started"
exit 0
}
nohup $DAEMON $DAEMONOPTS >> /dev/null 2>&1 &
;;
stop)
killall $DAEMON_BASE
;;
info)
echo "Name: webserver"
;;
status)
status
exit $?
;;
*)
echo "unsupported action: $1" >&2
exit 1
;;
esac
Rendez le script exécutable :
[root@node2 init.d]# chmod u+x weblauncher
Testez le script :
[root@node2 init.d]# ./weblauncher status [root@node2 init.d]# echo $? 1 [root@node2 init.d]# ./weblauncher start [root@node2 init.d]# ./weblauncher status [root@node2 init.d]# echo $? 0 [root@node2 init.d]# ./weblauncher stop [root@node2 init.d]# ./weblauncher status [root@node2 init.d]# echo $? 1
Arrêtez maintenant le service svc1 sur node1 :
[root@node1 ~]# svcmgr -s svc1 stop node1.opensvc.loc.svc1 stop action requested
Démarrez le service sur node2 :
[root@node2 ~]# svc1 start --local node2.opensvc.loc.svc1.ip#0 checking 192.168.121.42 availability node2.opensvc.loc.svc1.ip#0 ifconfig bond0:1 192.168.121.42 netmask 255.255.255.0 up node2.opensvc.loc.svc1.ip#0 send gratuitous arp to announce 192.168.121.42 is at bond0 node2.opensvc.loc.svc1.disk#0 vg vgsvc1 is already up node2.opensvc.loc.svc1.fs#app ext4 /dev/mapper/vgsvc1-lvappsvc1@/svc1/app is already mounted node2.opensvc.loc.svc1.fs#data ext4 /dev/mapper/vgsvc1-lvdatasvc1@/svc1/data is already mounted node2.opensvc.loc.svc1.app#web exec /svc1/app/init.d/weblauncher start as user root node2.opensvc.loc.svc1.app#web start done in 0:00:01.008940 ret 0
Haute Disponibilité
Afin de mettre en place la haute disponibilité, il convient de changer le mode d'orchestration à ha. Editez donc le fichier /etc/opensvc/svc1.conf sur node2 en ajoutant orchestrate = ha dans la section [DEFAULT] et monitor = True dans la section [app#web] :
[root@node2 ~]# svc1 edit config [root@node2 ~]# cat /etc/opensvc/svc1.conf [DEFAULT] id = 7adc8d25-1668-4078-a3f0-5b4472c436d2 app = MyApp nodes = node1.opensvc.loc node2.opensvc.loc orchestrate = ha [ip#0] ipname = svc1.opensvc.loc ipdev = bond0 [disk#0] type = vg name = vgsvc1 pvs = /dev/loop0 [fs#app] type = ext4 dev = /dev/mapper/vgsvc1-lvappsvc1 mnt = /svc1/app [fs#data] type = ext4 dev = /dev/mapper/vgsvc1-lvdatasvc1 mnt = /svc1/data [app#web] monitor = True script = weblauncher start = 10 check = 10 stop = 90
Siasissez la commande suivante puis testez la HA :
[root@node2 ~]# svc1 thaw
<html> <DIV ALIGN=“CENTER”> Copyright © 2020 Hugh Norris. </DIV> </html>
