OceanBaseデータベース

SQL - V4.4.2

PDF をダウンロード

このページには現在中国語/日本語版がなく、英語版を表示しています。

(Optional) Perform disk planning

最終更新:2026-09-04 11:28:25

The running of the OceanBase Database servers relies on the data disk, the clog disk, and the installation disk of OceanBase Database. If you are an individual user, you can put all the data on a single disk and skip this topic. If you are an enterprise user, we recommend that you separately mount your data to three disks.

If you do not have three disks for a server, or if you are using a Redundant Array of Independent Disks (RAID), you need to partition the disk or the logical volume of the disk array. We recommend that you use the following partitioning scheme:

Data disk

The data disk is used to store baseline data, and its path is specified by the data_dir parameter. When you start the OceanBase cluster for the first time, ${data_dir}/{sstable,slog} will be created automatically. The size of the data disk is determined by the datafile_disk_percentage/datafile_size parameters. Additionally, after deployment, you can perform dynamic expansion of disk files by using the datafile_next and datafile_maxsize parameters. For more information, see Configure automatic scale-out of disk space for data files.

Note

The current version of OceanBase Database supports separating slog from the data disk, meaning that slog and data files do not need to be on the same disk. For more information about the installation directory of OceanBase Database, see Installation directory structure of an OBServer node.
Clog disk

The path of the clog disk is specified by the clog_dir parameter. We recommend that you set the size of the clog disk to three to four times or more of the memory size of OceanBase Database. When you start the OceanBase cluster for the first time, ${clog_dir} will be created automatically. The clog disk contains multiple fixed-size files, and the clog files can be automatically created and cleared based on your needs. Automatic clearance is triggered when the clog reaches 80% of the total disk capacity. However, the clog files can be deleted only when the corresponding memory data has been merged into the baseline data.

For a given amount of data, the size of the clog is approximately three times the size of the memory data. Therefore, the maximum space required for the clog disk is directly proportional to the total amount of data after two major compactions. Empirical formula: clog file size = 3 to 4 times the upper limit of incremental data memory.
Installation disk of OceanBase Database

The path of the installation disk of OceanBase Database is specified by the home_path parameter. We recommend that you reserve at least 200 GB of space for the installation disk of OceanBase Database to save logs for seven days or more. The RPM package installation directory of OceanBase Database is located under ${home_path}. The baseline data files and clog files are mapped to the independent data disk and clog disk through soft links. The operational logs of OceanBase Database are located under ${home_path}/log. The operational logs will keep growing, and OceanBase Database cannot automatically delete the operational logs. Therefore, you need to regularly delete the operational logs.

Mount disks

Mounting requirements

The disk mounting requirements of an OCP server are listed in the following table.

Mount point	Size	Purpose	File system format
/home	100 GB to 300 GB	The operational log disk of each component	ext4 or xfs is recommended
/data/log1	3 to 4 times the memory size allocated to observer	The clog disk of the OCP meta database	ext4 or xfs is recommended
/data/1	Depends on the size of the data to store	The data disk of the OCP meta database	ext4 or xfs is recommended

The disk mounting requirements of an OBServer node are listed in the following table.

Directory	Size	Purpose	File system format
/home	100 GB to 300 GB	The installation disk of observer	ext4 or xfs is recommended
/data/log1	3 to 4 times the memory size allocated to observer	The log disk of the observer process	ext4 or xfs is recommended
/data/1	Depends on the size of the data to store	The data disk of the observer process	ext4 or xfs is recommended

Note

We recommend that the root directory be at least 50 GB. If you use LVM, we recommend that you use the striping parameter when creating a volume. Example: lvcreate -n data -L 3000G obvg --stripes=3 --stripesize=128.
In the production environment, we recommend that you use different disks for the data disk, the log disk, and the installation disk to avoid performance issues.

Mounting operation

Mounting disks must be performed under the root user. There are two ways to mount disks. You can choose an appropriate one based on your actual situation:

Mount disks by using the LVM tool (recommended).
Mount disks by using the fdisk tool.

Mount disks by using the LVM tool

Mount disks by using the fdisk tool

View the disk information

Use the fdisk -l command to identify the available disks and partitions, and confirm the target device (such as /dev/sdb1).
```
[root@test001 ~]# fdisk -l
```
(Optional) Install the LVM tool

If LVM is not preinstalled, run the following command to install LVM. If LVM has been installed, skip this step.
- Debian/Ubuntu
```
[root@test001 ~]# apt-get install lvm2
```
- CentOS/RHEL
```
[root@test001 ~]# yum install lvm2
```
Create a physical volume (PV)

Initialize the partition as a physical volume.

Notice

When a partition is initialized as a physical volume, the partition is reformatted, and the data will be lost. Proceed with caution.
```
[root@test001 ~]# pvcreate /dev/sdb1
```
After the creation, you can run the pvs command to verify the physical volume creation result.
Create a volume group (VG)

Merge multiple physical volumes into a VG.
```
[root@test001 ~]# vgcreate vg01 /dev/sdb1 /dev/sdc1 
```
After the merge, you can run the vgs command to view the volume group information.
Create a logical volume (LV)

Divide a logical volume of 100 GB from the VG. This is only an example, and the size of the logical volume can be set based on the actual situation.
```
[root@test001 ~]# lvcreate -L 100G -n lv01 vg01
```
After the creation, you can run the lvs command to view the logical volume information.

Format and mount

Format the logical volume as an ext4 file system.
```
[root@test001 ~]# mkfs.ext4 /dev/vg01/lv01
```
Create a mount point.
```
[root@test001 ~]# mkdir -p /data/1
```

Mount the logical volume temporarily.

[root@test001 ~]# mount /dev/vg01/lv01 /data/1

Set automatic mounting on startup

Edit the /etc/fstab file and add the mounting configuration:
```
[root@test001 ~]# vim /etc/fstab
```
Add the following content to the /etc/fstab file:
```
/dev/vg01/lv01  /data/1  ext4  defaults,noatime,nodiratime,nodelalloc,barrier=0  0  0
```

View the disk information

Use the fdisk -l command to identify the available disks and partitions, and confirm the target device (such as /dev/sdb1).
```
[root@test001 ~]# fdisk -l
```
Create a partition

Use the fdisk tool to create a new partition. For example, run fdisk /dev/sdb1, enter n to create a primary partition, and finally save the changes (w).
```
[root@test001 ~]# fdisk /dev/sdb1
```

Format and mount

Format the partition as an ext4 file system.
```
[root@test001 ~]# mkfs.ext4 /dev/sdb1
```
Create a mount point.
```
[root@test001 ~]# mkdir -p /data/1
```

Mount the partition temporarily.

[root@test001 ~]# mount /dev/sdb1 /data/1

Set automatic mounting on startup.

Edit the /etc/fstab file and add the mounting configuration:
```
[root@test001 ~]# vim /etc/fstab
```
Add the following content to the /etc/fstab file:
```
/dev/sdb1  /data/1  ext4  defaults,noatime,nodiratime,nodelalloc,barrier=0  0  0
```

View the disks

After the disks are mounted, run the following command to check the disk partitioning:

[root@test001 ~]# df -h

The returned result is as follows:

Filesystem      Size  Used Avail Use% Mounted on
devtmpfs         31G     0   31G   0% /dev
tmpfs            31G     0   31G   0% /dev/shm
tmpfs            31G  516K   31G   1% /run
tmpfs            31G     0   31G   0% /sys/fs/cgroup
/dev/vda1       493G  171G  302G  37% /
tmpfs           6.2G     0  6.2G   0% /run/user/0
/dev/sdb1     984G    77M  934G  1%  /data/1
/dev/vdc1     196G     61M   186G  1%  /data/log1
/dev/vdb1     492G    73M   467G  1%  /home/admin/oceanbase

Result description:

/data/1 is the data disk, which is 1 TB in size.
/data/log1 is used to store logs.
/home/admin/oceanbase is used to store the binary files and operational logs of OceanBase Database.

Make sure that the disks corresponding to data_dir, redo_dir, and home_path in the configuration file have been mounted. The directories corresponding to data_dir and redo_dir must be empty, and the disk usage of the directory corresponding to data_dir must be lower than 4%.

Set directory permissions

After the disks are mounted, you need to check the permissions of the directories corresponding to the disk mounting. You can run the following command to check the directory permissions of the cluster-related files.

Taking the data directory as an example:

[root@test001 data]# ls -al

The returned result is as follows, indicating that the owner of the directory is the admin user in the admin group:

drwxr-xr-x 2 admin admin 4096 02/09 18:43 .
drwxr-xr-x 2 admin admin 4096 02/09 18:43 log1

If you find that the admin user does not have the permissions on the related files, you can run the following commands to change the file owner:

[root@test001 ~]# chown -R admin:admin /data/log1
[root@test001 ~]# chown -R admin:admin /data

Here, /data/log1 and /data are example mount directories. You need to replace them with your actual mount directories.

次のトピック

What's New

Note

Mount disks

Mounting requirements

Note

Mounting operation

Notice

View the disks

Set directory permissions