ビルド済みパッケージからインストール

第3章コンパイル済みDRBDバイナリパッケージのインストール
- ubuntu16.04以降、debian8用

/etc/apt/sources.list.d/drbd9.list

deb http://ppa.launchpad.net/linbit/linbit-drbd9-stack/ubuntu xenial main
# deb http://ppa.launchpad.net/linbit/linbit-drbd9-stack/ubuntu trusty main
# deb http://ppa.launchpad.net/linbit/linbit-drbd9-stack/ubuntu devel main

apt-key adv --keyserver keyserver.ubuntu.com --recv-keys 34893610CEAA9512
apt update
apt-get install -y drbd-utils python-drbdmanage drbd-dkms

/lib/modules/4.8.0-39-generic/updates/dkms/
- drbd.ko
- drbd_transport_tcp.ko

Debian8ではtrustyのリポジトリを使う。

xenialのリポジトリだと以下エラーになる。

drbd-utils : Depends: libstdc++6 (>= 5.2) but 4.9.2-10 is to be installed

echo "deb http://ppa.launchpad.net/linbit/linbit-drbd9-stack/ubuntu trusty main" > /etc/apt/sources.list.d/drbd9.list
apt-key adv --keyserver keyserver.ubuntu.com --recv-keys 34893610CEAA9512
apt update
apt-get install -y drbd-utils python-drbdmanage drbd-dkms

ソースからインストール

apt install -y "build-essential git checkinstall"
# hostname build

drbd-9.0

git clone "git://git.drbd.org/drbd-9.0.git"
cd drbd-9.0
# export KDIR="/lib/modules/4.8.0-38-generic/build"
make -s
make install
checkinstall -y -D --pkgname="drbd-9.0"

drbd-9.0_9.0-1_amd64.debが出来上がる

drbd-utils

apt install -y "autoconf pkg-config flex xsltproc"
git clone "git://git.drbd.org/drbd-utils.git"
cd drbd-utils
./autogen.sh
./configure --prefix=/usr --localstatedir=/var --sysconfdir=/etc

# 時間が20分はかかる
make -s
make install
checkinstall -y -D --pkgname="drbd-utils"

drbdmanage

git clone "git://git.drbd.org/drbdmanage.git"
cd drbdmanage
make -s
make install
checkinstall -y -D --pkgname="drbdmanage"

準備 : CloudAtCostの場合

CloudAtCost の場合、 Debian8で15GBのDISKを選ぶと以下構成になる。

  Device     Boot    Start      End  Sectors  Size Id Type
  /dev/sda1  *        2048   499711   497664  243M 83 Linux
  /dev/sda2         501758 20969471 20467714  9.8G  5 Extended
+ /dev/sda3       20969472 30719999  9750528  4.7G 83 Linux
  /dev/sda5         501760 20969471 20467712  9.8G 8e Linux LVM

/dev/sda5 が 10GBで作成、
/dev/sda3 に残りのDISK容量割当てられる。
- このsda3を領域として使う。

Debian8(jessie) にて確認。trustyリポジトリでインストールしておく。

ホスト名	ip	役割
drbd1	xxx.xxx.xxx.101	master
drbd2	xxx.xxx.xxx.102	slave
drbd3	xxx.xxx.xxx.103	slave

drbdmanage init

drbdpool という Volume group が必要。

確認

  $ drbdmanage init
  
  You are going to initialize a new drbdmanage cluster.
  CAUTION! Note that:
    * Any previous drbdmanage cluster information may be removed
    * Any remaining resources managed by a previous drbdmanage installation
      that still exist on this system will no longer be managed by drbdmanage
  
  Confirm:
  
    yes/no: yes
  Error: External command failed:
  lvcreate -n .drbdctrl_0 -L 4m drbdpool
  Command output:
+   (stderr)   Volume group "drbdpool" not found
  Command exited with exit_code 5
  
  Initialization failed

drbdpool作成

vgcreate drbdpool /dev/sda3

drbdpool作成やり直したい場合

# dd if=/dev/zero of=/dev/sda3 bs=1M count=100
# reboot

init

$ drbdmanage init  # 1台目でのみ実施

You are going to initialize a new drbdmanage cluster.
CAUTION! Note that:
  * Any previous drbdmanage cluster information may be removed
  * Any remaining resources managed by a previous drbdmanage installation
    that still exist on this system will no longer be managed by drbdmanage

Confirm:

  yes/no: yes
Empty drbdmanage control volume initialized on '/dev/drbd0'.
Empty drbdmanage control volume initialized on '/dev/drbd1'.
Waiting for server: .
Operation completed successfully

# 間違って2台目、3台目でも実行してしまった場合は、以下コマンドで戻せる。
# drbdmanage uninit

確認

確認

$ drbdadm status
.drbdctrl role:Primary
  volume:0 disk:UpToDate
  volume:1 disk:UpToDate

/etc/drbd.d/drbdctrl.res が自動で作成された

/etc/drbd.d/drbdctrl.res

resource .drbdctrl {
    net {
        cram-hmac-alg       sha256;
        shared-secret       "xxxxxxxxxxxxxxxxxxxx";
        allow-two-primaries no;
    }
    volume 0 {
        device      minor 0;
        disk        /dev/drbdpool/.drbdctrl_0;
        meta-disk   internal;
    }
    volume 1 {
        device      minor 1;
        disk        /dev/drbdpool/.drbdctrl_1;
        meta-disk   internal;
    }
    on drbd1 {
        node-id     0;
        address     ipv4 xxx.xxx.xxx.101:6999;
    }
    connection-mesh {
        hosts drbd1;
        net {
            protocol C;
        }
    }
}

poolサイズ(4752MBの未使用領域)

$ drbdmanage list-nodes
+--------------------------------------------------------------+
| Name      | Pool Size | Pool Free |                  | State |
|--------------------------------------------------------------|
| localhost |      4760 |      4752 |                  |    ok |
+--------------------------------------------------------------+

add nodes

drbdmanage add-node drbd2 xxx.xxx.xxx.102
drbdmanage add-node drbd3 xxx.xxx.xxx.103

確認

$ drbdmanage list-nodes
+----------------------------------------------------------------+
| Name  | Pool Size | Pool Free |                        | State |
|----------------------------------------------------------------|
| drbd1 |      4760 |      4752 |                        |    ok |
| drbd2 |      4760 |      4752 |                        |    ok |
| drbd3 |      4760 |      4752 |                        |    ok |
+----------------------------------------------------------------+

status

$ drbdadm status
.drbdctrl role:Primary
  volume:0 disk:UpToDate
  volume:1 disk:UpToDate
  drbd2 role:Secondary
    volume:0 peer-disk:UpToDate
    volume:1 peer-disk:UpToDate
  drbd3 role:Secondary
    volume:0 peer-disk:UpToDate
    volume:1 peer-disk:UpToDate

/etc/drbd.d/drbdctrl.res

  resource .drbdctrl {
      net {
          cram-hmac-alg       sha256;
          shared-secret       "xxxxxxxxxxxxxxxxxxxx";
          allow-two-primaries no;
      }
      volume 0 {
          device      minor 0;
          disk        /dev/drbdpool/.drbdctrl_0;
          meta-disk   internal;
      }
      volume 1 {
          device      minor 1;
          disk        /dev/drbdpool/.drbdctrl_1;
          meta-disk   internal;
      }
      on drbd1 {
          node-id     0;
          address     ipv4 xxx.xxx.xxx.101:6999;
      }
+     on drbd3 {
+         node-id     2;
+         address     ipv4 xxx.xxx.xxx.102:6999;
+     }
+     on drbd2 {
+         node-id     1;
+         address     ipv4 xxx.xxx.xxx.103:6999;
+     }
      connection-mesh {
+         hosts drbd1 drbd3 drbd2;
          net {
              protocol C;
          }
      }
  }

ノードの削除

エラーの場合

$ drbdmanage list-nodes
+------------------------------------------------------------------------------------------------------------+
| Name  | Pool Size | Pool Free |         |                                                            State |
|------------------------------------------------------------------------------------------------------------|
| drbd1 |      4760 |      4752 |         |                                                               ok |
| drbd2 |   unknown |   unknown |         | offline/quorum vote ignored, pending actions: adjust connections |
| drbd3 |   unknown |   unknown |         | offline/quorum vote ignored, pending actions: adjust connections |
+------------------------------------------------------------------------------------------------------------+

複数同時に書ける

drbdmanage delete-node drbd2 drbd3

リソース登録

master機でのみ実施

drbdmanage add-resource r0
# 削除は drbdmanage delete-resource r0

$ drbdmanage list-volumes --groupby Size --show Port
+------------------------------------------------------------------+
| Name | Vol ID | Size | Minor | Port |                    | State |
|------------------------------------------------------------------|
| r0   |      * |    * |     * | 7000 |                    |     * |
+------------------------------------------------------------------+

r0に4Gのvolume作成

drbdmanage add-volume r0 4

削除

指定を間違えた場合

# drbdmanage list-volumes --groupby Size --show Port
+--------------------------------------------------------------+
| Name | Vol ID |     Size | Minor | Port |            | State |
|--------------------------------------------------------------|
| r0   |      1 |    4 GiB |   101 | 7000 |            |    ok |
| r0   |      0 | 4.65 TiB |   100 | 7000 |            |    ok |
+--------------------------------------------------------------+

root@drbd1:~# drbdmanage delete-volume r0 
usage: drbdmanage remove-volume [-h] [-q] [-f] name vol_id
drbdmanage remove-volume: error: too few arguments

root@drbd1:~# drbdmanage delete-volume r0 0
You are going to remove a volume from all nodes of the cluster.
Please confirm:
  yes/no: yes
Operation completed successfully

root@drbd1:~# drbdmanage delete-volume r0 1
You are going to remove a volume from all nodes of the cluster.
Please confirm:
  yes/no: yes
Operation completed successfully

デプロイ

3台に配布

drbdmanage deploy-resource r0 3

失敗事例

mkfs.ext4 /dev/drbd0

root@drbd1:~# drbdmanage add-resource r0
Error: I/O error while accessing persistent configuration storage

root@drbd1:~# drbdmanage list-volumes --groupby Size --show Port
+-----------------------------------------------------------------+
| Name | Vol ID | Size | Minor | Port |                   | State |
|-----------------------------------------------------------------|
| r0   |      * |    * |     * | 7000 |                   |     * |
+-----------------------------------------------------------------+

やり直し。