[Subscription] Flush lingering batch before WAL gap retry
IoTDB (Internet of Things Database) is a data management system for time series data, which provides users with specific services, including data collection, storage and analysis. Due to its lightweight structure, high performance and usable features, together with its seamless integration with the Hadoop and Spark ecosystem, IoTDB meets the requirements of massive dataset storage, high throughput data input, and complex data analysis in the industrial IoT field.
IoTDB depends on TsFile which is a columnar storage file format designed for time series data. The branch iotdb of TsFile project is used to deploy SNAPSHOT version for IoTDB project.
The main features of IoTDB are as follows:
For the latest information about IoTDB, please visit IoTDB official website.
This short guide will walk you through the basic process of using IoTDB. For a more detailed introduction, please visit our website's User Guide.
To use IoTDB, you need to have:
# Linux > sudo sysctl -w net.core.somaxconn=65535 # FreeBSD or Darwin > sudo sysctl -w kern.ipc.somaxconn=65535
⚠️ Important: System Resource Limits
IoTDB requires sufficient system resource limits to start correctly. If these limits are not configured, IoTDB may fail to start or emit warnings such as
"too many open files"or"connection reset"in the logs.Linux / macOS
ulimit -n 65535 sudo sysctl -w net.core.somaxconn=65535Docker Users When running IoTDB in Docker, these limits must be applied on the host machine or explicitly passed to the container:
docker run --ulimit nofile=65535:65535 ...Windows Users These commands are not applicable on Windows. Please follow the Windows installation steps below and ensure sufficient system resources.
IoTDB provides two installation methods, you can refer to the following suggestions, choose the one fits you best:
Please contact timechodb staff to obtain the IoTDB installation package. This method provides pre-compiled binary executables, ready to use out of the box, suitable for quick deployment and direct use.
Please contact timechodb staff to obtain the Docker image. This method supports containerized deployment and is suitable for scenarios requiring flexible environment configuration or rapid migration.
For more specific installation methods, please visit our website's User Guide.
Skip this chapter if you are using Windows.
As we use Thrift for our RPC module (communication and protocol definition), we involve Thrift during the compilation, so Thrift compiler 0.13.0 (or higher) is required to generate Thrift Java code. Thrift officially provides binary compiler for Windows, but unfortunately, they do not provide that for Unix OSs.
If you have permission to install new software, use apt install or yum install or brew install to install the Thrift compiler. (If you already have installed the thrift compiler, skip this step.) Then, you may add the following parameter when running Maven: -Dthrift.download-url=http://apache.org/licenses/LICENSE-2.0.txt -Dthrift.exec.absolute.path=<YOUR LOCAL THRIFT BINARY FILE>.
If not, then you have to compile the thrift compiler, and it requires you install a boost library first. Therefore, we compiled a Unix compiler ourselves and put it onto GitHub, and with the help of a maven plugin, it will be downloaded automatically during compilation. This compiler works fine with gcc8 or later, Ubuntu MacOS, and CentOS, but previous versions and other OSs are not guaranteed.
If you can not download the thrift compiler automatically because of a network problem, you can download it by yourself, and then either: rename your thrift file to {project_root}\thrift\target\tools\thrift_0.12.0_0.13.0_linux.exe; or, add Maven commands: -Dthrift.download-url=http://apache.org/licenses/LICENSE-2.0.txt -Dthrift.exec.absolute.path=<YOUR LOCAL THRIFT BINARY FILE>.
You can download the source code from:
git clone https://github.com/apache/iotdb.git
The default dev branch is the master branch, if you want to use a released version x.x.x:
git checkout vx.x.x
Or checkout to the branch of a big version, e.g., the branch of 1.0 is rel/1.0.
git checkout rel/x.x
Under the root path of iotdb:
> mvn clean package -pl distribution -am -DskipTests
After being built, the IoTDB distribution is located at the folder: “distribution/target”.
The distribution build also produces apache-iotdb-<version>-edge-bin.zip. IoTDB Edge runs the ConfigNode and DataNode in one JVM for resource-constrained, single-node deployments. It is an additional artifact and does not replace any existing distribution package.
After extracting the package, configure conf/iotdb-system.properties, then start or stop the Edge process with:
sbin/start-edge.sh sbin/stop-edge.sh
On Windows, use sbin\windows\start-edge.bat and sbin\windows\stop-edge.bat. The package retains tools that are compatible with the combined Edge process.
Under the iotdb/iotdb-client path:
> mvn clean package -pl cli -am -DskipTests
After being built, the IoTDB cli is located at the folder “cli/target”.
IoTDB supports compile-time internationalization (i18n) for log and error messages. By default, messages are in English. To build with Chinese messages, activate the with-zh-locale Maven profile:
> mvn clean package -pl distribution -am -DskipTests -P with-zh-locale
This works by swapping the source directory src/main/i18n/en (default) with src/main/i18n/zh, where each module keeps locale-specific Java constant classes containing translated message strings.
Use -P with-cpp for compiling the cpp client. (For more details, read client-cpp's Readme file.)
NOTE: Directories “thrift/target/generated-sources/thrift”, “thrift-sync/target/generated-sources/thrift”, “thrift-cluster/target/generated-sources/thrift”, “thrift-influxdb/target/generated-sources/thrift” and “antlr/target/generated-sources/antlr4” need to be added to sources roots to avoid compilation errors in the IDE.
In IDEA, you just need to right click on the root project name and choose “Maven->Reload Project” after you run mvn package successfully.
Configuration files are under the “conf” folder.
datanode-env.bat, datanode-env.sh),iotdb-system.properties)logback.xml).For more information, please see Config Manual.
You can go through the following steps to test the installation. If there is no error returned after execution, the installation is completed.
Users can start 1C1D IoTDB by the start-standalone script under the sbin folder.
# Unix/OS X > sbin/start-standalone.sh # Windows > sbin\start-standalone.bat
IoTDB offers different ways to interact with server, here we introduce the basic steps of using Cli tool to insert and query data.
After installing IoTDB, there is a default user ‘root’, its default password is ‘TimechoDB@2021’. Users can use this default user to login Cli to use IoTDB. The start-up script of Cli is the start-cli script in the folder sbin. When executing the script, user should assign IP, PORT, USER_NAME and PASSWORD. The default parameters are “-h 127.0.0.1 -p 6667 -u root -pw TimechoDB@2021”.
Here is the command for starting the Cli:
# Unix/OS X > sbin/start-cli.sh -h 127.0.0.1 -p 6667 -u root -pw root # Windows > sbin\start-cli.bat -h 127.0.0.1 -p 6667 -u root -pw root
The command line cli is interactive, so you should see the welcome logo and statements if everything is ready:
_____ _________ ______ ______ |_ _| | _ _ ||_ _ `.|_ _ \ | | .--.|_/ | | \_| | | `. \ | |_) | | | / .'`\ \ | | | | | | | __'. _| |_| \__. | _| |_ _| |_.' /_| |__) | |_____|'.__.' |_____| |______.'|_______/ version x.x.x IoTDB> login successfully IoTDB>
Now, let us introduce the way of creating timeseries, inserting data and querying data.
The data in IoTDB is organized as timeseries. Each timeseries includes multiple data–time pairs, and is owned by a database. Before defining a timeseries, we should define a database using CREATE DATABASE first, and here is an example:
IoTDB> CREATE DATABASE root.ln
We can also use SHOW DATABASES to check the database being created:
IoTDB> SHOW DATABASES +-------------+ | Database| +-------------+ | root.ln| +-------------+ Total line number = 1
After the database is set, we can use CREATE TIMESERIES to create a new timeseries. When creating a timeseries, we should define its data type and the encoding scheme. Here we create two timeseries:
IoTDB> CREATE TIMESERIES root.ln.wf01.wt01.status WITH DATATYPE=BOOLEAN, ENCODING=PLAIN IoTDB> CREATE TIMESERIES root.ln.wf01.wt01.temperature WITH DATATYPE=FLOAT, ENCODING=RLE
In order to query the specific timeseries, we can use SHOW TIMESERIES . represent the location of the timeseries. The default value is “null”, which queries all the timeseries in the system (the same as using “SHOW TIMESERIES root”). Here are some examples:
IoTDB> SHOW TIMESERIES +-----------------------------+-----+-------------+--------+--------+-----------+----+----------+ | Timeseries|Alias|Database|DataType|Encoding|Compression|Tags|Attributes| +-----------------------------+-----+-------------+--------+--------+-----------+----+----------+ |root.ln.wf01.wt01.temperature| null| root.ln| FLOAT| RLE| SNAPPY|null| null| | root.ln.wf01.wt01.status| null| root.ln| BOOLEAN| PLAIN| SNAPPY|null| null| +-----------------------------+-----+-------------+--------+--------+-----------+----+----------+ Total line number = 2
IoTDB> SHOW TIMESERIES root.ln.wf01.wt01.status +------------------------+-----+-------------+--------+--------+-----------+----+----------+ | timeseries|alias|database|dataType|encoding|compression|tags|attributes| +------------------------+-----+-------------+--------+--------+-----------+----+----------+ |root.ln.wf01.wt01.status| null| root.ln| BOOLEAN| PLAIN| SNAPPY|null| null| +------------------------+-----+-------------+--------+--------+-----------+----+----------+ Total line number = 1
Inserting timeseries data is a basic operation of IoTDB, you can use the ‘INSERT’ command to finish this. Before insertion, you should assign the timestamp and the suffix path name:
IoTDB> INSERT INTO root.ln.wf01.wt01(timestamp,status) values(100,true); IoTDB> INSERT INTO root.ln.wf01.wt01(timestamp,status,temperature) values(200,false,20.71)
The data that you have just inserted will be displayed as follows:
IoTDB> SELECT status FROM root.ln.wf01.wt01 +------------------------+------------------------+ | Time|root.ln.wf01.wt01.status| +------------------------+------------------------+ |1970-01-01T00:00:00.100Z| true| |1970-01-01T00:00:00.200Z| false| +------------------------+------------------------+ Total line number = 2
You can also query several timeseries data using one SQL statement:
IoTDB> SELECT * FROM root.ln.wf01.wt01 +------------------------+-----------------------------+------------------------+ | Time|root.ln.wf01.wt01.temperature|root.ln.wf01.wt01.status| +------------------------+-----------------------------+------------------------+ |1970-01-01T00:00:00.100Z| null| true| |1970-01-01T00:00:00.200Z| 20.71| false| +------------------------+-----------------------------+------------------------+ Total line number = 2
To change the time zone in Cli, you can use the following SQL:
IoTDB> SET time_zone=+08:00 Time zone has set to +08:00 IoTDB> SHOW time_zone Current time zone: Asia/Shanghai
Add then the query result will show using the new time zone.
IoTDB> SELECT * FROM root.ln.wf01.wt01 +-----------------------------+-----------------------------+------------------------+ | Time|root.ln.wf01.wt01.temperature|root.ln.wf01.wt01.status| +-----------------------------+-----------------------------+------------------------+ |1970-01-01T08:00:00.100+08:00| null| true| |1970-01-01T08:00:00.200+08:00| 20.71| false| +-----------------------------+-----------------------------+------------------------+ Total line number = 2
The commands to exit the Cli are:
IoTDB> quit or IoTDB> exit
For more information about the commands supported by IoTDB SQL, please see User Guide.
The server can be stopped with “ctrl-C” or the following script:
# Unix/OS X > sbin/stop-standalone.sh # Windows > sbin\stop-standalone.bat
see The use of Data Import Tool see The use of Data Export Tool
see The use of CSV Import and Export Tool
see Frequent Questions when Compiling the Source Code
tietouqiao.