Environment Requirements

1. Disk Array

1.1 Configuration Suggestions

IoTDB does not have strict operational requirements for disk array configurations. It is recommended to use multiple disk arrays to store IoTDB data to achieve concurrent writing across multiple disk arrays. The following configuration suggestions can be referenced:

1. Physical Environment
   • System Disk: It is recommended to use 2 disks for RAID1, considering only the space occupied by the operating system itself. No additional space needs to be reserved for IoTDB on the system disk.
   • Data Disk:
     • It is recommended to use RAID for data protection at the disk level.
     • It is recommended to provide multiple disks (around 1-6) or disk groups for IoTDB (avoiding creating a single disk array with all disks, as it may affect IoTDB's performance ceiling).
2. Virtual Environment
   • It is recommended to mount multiple hard drives (around 1-6).

1.2 Configuration Examples

• Example 1: 4 x 3.5-inch Hard Drives

  • Since the server has fewer installed hard drives, RAID5 can be directly configured without additional settings.

  • Recommended configuration:

    Classification	RAID Type	Number of Hard Drives	Redundancy	Usable Drives
    System/Data Disk	RAID5	4	1 disk failure allowed	3

• Example 2: 12 x 3.5-inch Hard Drives

  • The server is configured with 12 x 3.5-inch hard drives.

  • The first 2 disks are recommended for RAID1 as the system disk. The data disks can be divided into 2 groups of RAID5, with 5 disks in each group (4 usable).

  • Recommended configuration:

    Classification	RAID Type	Number of Hard Drives	Redundancy	Usable Drives
    System Disk	RAID1	2	1 disk failure allowed	1
    Data Disk	RAID5	5	1 disk failure allowed	4
    Data Disk	RAID5	5	1 disk failure allowed	4

• Example 3: 24 x 2.5-inch Hard Drives

  • The server is configured with 24 x 2.5-inch hard drives.

  • The first 2 disks are recommended for RAID1 as the system disk. The remaining disks can be divided into 3 groups of RAID5, with 7 disks in each group (6 usable). The last disk can be left idle or used for storing write-ahead logs.

  • Recommended configuration:

    Usage Classification	RAID Type	Number of Hard Drives	Redundancy	Usable Drives
    System Disk	RAID1	2	1 disk failure allowed	1
    Data Disk	RAID5	7	1 disk failure allowed	6
    Data Disk	RAID5	7	1 disk failure allowed	6
    Data Disk	RAID5	7	1 disk failure allowed	6
    Data Disk	No RAID	1	Data loss if damaged	1

2. Operating System

2.1 Version Requirements

IoTDB supports operating systems such as Linux, Windows, and MacOS. TimechoDB also supports Chinese CPUs like Loongson, Phytium, and Kunpeng, as well as Chinese operating systems like Kylin, UOS, and NingSi.

2.2 Hard Disk Partitioning

• It is recommended to use the default standard partitioning method. LVM expansion and hard disk encryption are not recommended.
• The system disk only needs to meet the space requirements of the operating system. No additional space needs to be reserved for IoTDB.
• Each disk group should correspond to a single partition. Data disks (with multiple disk groups corresponding to RAID) do not need additional partitioning, and all space should be allocated to IoTDB.

Recommended disk partitioning is as follows:

2.3 Network Configuration

1. Disable Firewall

   # Check firewall status
   systemctl status firewalld
   # Stop firewall
   systemctl stop firewalld
   # Permanently disable firewall
   systemctl disable firewalld
   

2. Ensure Required Ports Are Not Occupied
   • Cluster Ports: By default, ConfigNode uses ports 10710 and 10720, while DataNode uses ports 6667, 10730, 10740, 10750, 10760, 9090, 9190, and 3000. Ensure these ports are not occupied. Check as follows:

     lsof -i:6667   or  netstat -tunp | grep 6667
     lsof -i:10710   or  netstat -tunp | grep 10710
     lsof -i:10720   or  netstat -tunp | grep 10720
     # If the command outputs anything, the port is occupied.
     

   • Cluster Deployment Tool Ports: When using the cluster management tool opskit for installation and deployment, ensure the SSH remote connection service is configured and port 22 is open.

     yum install openssh-server            # Install SSH service
     systemctl start sshd                  # Enable port 22
     

3. Ensure Network Connectivity Between Servers

2.4 Other Configurations

1. Disable System Swap Memory

   echo "vm.swappiness = 0" >> /etc/sysctl.conf
   # Execute both swapoff -a and swapon -a to transfer data from swap back to memory and clear swap data.
   # Do not omit the swappiness setting and only execute swapoff -a; otherwise, swap will automatically reopen after reboot, rendering the operation ineffective.
   swapoff -a && swapon -a
   # Apply the configuration without rebooting.
   sysctl -p
   # Check memory allocation; swap should be 0.
   free -m
   

2. Set System Maximum Open Files to 65535 to avoid “too many open files” errors.

   # Check current limit
   ulimit -n
   # Temporarily modify
   ulimit -n 65535
   # Permanently modify
   echo "* soft nofile 65535" >> /etc/security/limits.conf
   echo "* hard nofile 65535" >> /etc/security/limits.conf
   # After exiting the current terminal session, check; it should display 65535.
   ulimit -n
   

3. Software Dependencies

Install Java Runtime Environment, Java version >= 1.8. Ensure JDK environment variables are set. (For versions V1.3.2.2 and later, it is recommended to directly deploy JDK17. Older JDK versions may have performance issues in some scenarios, and DataNode may fail to stop.)

 # Example of installing JDK-17 on CentOS7:
 tar -zxvf jdk-17_linux-x64_bin.tar     # Extract JDK files
 vim ~/.bashrc                          # Configure JDK environment
 {
   export JAVA_HOME=/usr/lib/jvm/jdk-17.0.9
   export PATH=$JAVA_HOME/bin:$PATH
 }  # Add JDK environment variables
 source ~/.bashrc                       # Apply environment configuration
 java -version                          # Check JDK environment