This tutorial shows you how to build a Spring Boot web application with CockroachDB, using the Spring Data JDBC module for data access. The code for the example application is available for download from GitHub, along with identical examples that use JPA, jOOQ, and MyBatis for data access.
This tutorial assumes you are running under SERIALIZABLE isolation. Client-side retry handling is not necessary under READ COMMITTED isolation.
Step 1. Start CockroachDB
Choose whether to run a local cluster or a free CockroachDB Cloud cluster.
- If you haven't already, download the CockroachDB binary.
- Start a local, secure cluster.
Create a free trial cluster
- Create a CockroachDB Cloud account. If this is your first CockroachDB Cloud organization, it will be credited with $400 in free trial credits to get you started.
- On the Get Started page, click Create cluster.
- On the Select a plan page, select Standard.
- On the Cloud & Regions page, select a cloud provider (GCP or AWS).
- In the Regions section, select a region for the cluster. Refer to CockroachDB Cloud Regions for the regions where CockroachDB Standard clusters can be deployed. To create a multi-region cluster, click Add region and select additional regions.
- Click Next: Capacity.
On the Capacity page, keep the Provisioned capacity at the default value of 2 vCPUs.
Click Next: Finalize.
On the Finalize page, name your cluster. If an active free trial is listed in the right pane, you will not need to add a payment method, though you will need to do this by the end of the trial to maintain your organization's clusters.
Click Create cluster.
Your cluster will be created in a few seconds and the Create SQL user dialog will display.
Set up your cluster connection
Once your cluster is created, the Connect to cluster-name dialog displays. Use the information provided in the dialog to set up your cluster connection for the SQL user that was created by default:
In your terminal, run the second command from the dialog to create a new
certsdirectory on your local machine and download the CA certificate to that directory:curl --create-dirs -o ~/.postgresql/root.crt -O https://cockroachlabs.cloud/clusters/{cluster-id}/certYour
certfile will be downloaded to~/.postgresql/root.crt.curl --create-dirs -o ~/.postgresql/root.crt -O https://cockroachlabs.cloud/clusters/{cluster-id}/certYour
certfile will be downloaded to~/.postgresql/root.crt.mkdir -p $env:appdata\.postgresql\; Invoke-WebRequest -Uri https://cockroachlabs.cloud/clusters/{cluster-id}/cert -OutFile $env:appdata\.postgresql\root.crtYour
certfile will be downloaded to%APPDATA%/.postgresql/root.crt.Copy the connection string provided, which will be used in the next steps (and to connect to your cluster in the future).
Warning:This connection string contains your password, which will be provided only once. If you forget your password, you can reset it by going to the SQL Users page for the cluster, found at
https://cockroachlabs.cloud/cluster/<CLUSTER ID>/users.cockroach sql --url 'postgresql://<username>:<password>@<cluster-host>:26257/defaultdb?sslmode=verify-full&sslrootcert='$HOME'/.postgresql/root.crt'cockroach sql --url 'postgresql://<username>:<password>@<cluster-host>:26257/defaultdb?sslmode=verify-full&sslrootcert='$HOME'/.postgresql/root.crt'cockroach sql --url "postgresql://<username>:<password>@<cluster-host>:26257/defaultdb?sslmode=verify-full&sslrootcert=$env:appdata/.postgresql/root.crt"Where:
<username>is the SQL user. By default, this is your CockroachDB Cloud account username.<password>is the password for the SQL user. The password will be shown only once in the Connection info dialog after creating the cluster.<cluster-hostname>is the hostname of your CockroachDB Cloud cluster.<cluster-id>is a unique string used to identify your cluster when downloading the CA certificate. For example,12a3bcde-4fa5-6789-1234-56bc7890d123.
You can find these settings in the Connection parameters tab of the Connection info dialog.
Step 2. Create a database and a user
Open a SQL shell to your local cluster using the
cockroach sqlcommand:$ cockroach sql --certs-dir={certs-dir} --host=localhost:{port}Where
{certs_dir}is the full path to thecertsdirectory that you created when setting up the cluster, and{port}is the port at which the cluster is listening for incoming connections.In the SQL shell, create the
roach_datadatabase that your application will use:> CREATE DATABASE roach_data;Create a SQL user for your app:
> CREATE USER {username} WITH PASSWORD {password};Take note of the username and password. You will use it to connect to the database later.
Give the user the necessary permissions:
> GRANT ALL ON DATABASE roach_data TO {username};Exit the shell, and generate a certificate and key for your user by running the following command:
$ cockroach cert create-client {user} --certs-dir={certs-dir} --ca-key={certs-dir}/ca.key --also-generate-pkcs8-key
The --also-generate-pkcs8-key flag generates a key in PKCS#8 format, which is the standard key encoding format in Java. In this case, the generated PKCS8 key will be named client.{user}.key.pk8.
- If you haven't already, download the CockroachDB binary.
Start the built-in SQL shell using the connection string you got from the CockroachDB Cloud Console earlier:
$ cockroach sql \ --url 'postgresql://<user>@<cluster-name>-<short-id>.<region>.<host>:26257/<database>?sslmode=verify-full&sslrootcert='$HOME'/Library/CockroachCloud/certs/<cluster-name>-ca.crt'Enter your SQL user password.
In the SQL shell, create the
roach_datadatabase that your application will use:> CREATE DATABASE roach_data;
Step 3. Install JDK
Download and install a Java Development Kit. Spring Boot supports Java versions 8, 11, and 14. In this tutorial, we use JDK 8 from OpenJDK.
Step 4. Install Maven
This example application uses Maven to manage all application dependencies. Spring supports Maven versions 3.2 and later.
To install Maven on macOS, run the following command:
$ brew install maven
To install Maven on a Debian-based Linux distribution like Ubuntu:
$ apt-get install maven
To install Maven on a Red Hat-based Linux distribution like Fedora:
$ dnf install maven
For other ways to install Maven, see its official documentation.
Step 5. Get the application code
To get the application code, download or clone the roach-data repository. The code for the example JDBC application is located under the roach-data-jdbc directory.
(Optional) To recreate the application project structure with the same dependencies as those used by this sample application, you can use Spring initializr with the following settings:
Project
- Maven Project
Language
- Java
Spring Boot
- 2.2.6
Project Metadata
- Group: io.roach
- Artifact: data
- Name: data
- Package name: io.roach.data
- Packaging: Jar
- Java: 8
Dependencies
- Spring Web
- Spring Data JDBC
- Spring Boot Actuator
- Spring HATEOS
- Liquibase Migration
- PostgreSQL Driver
Step 6. Run the application
Compiling and running the application code will start a web application, initialize the accounts table in the roach_data database, and submit some requests to the app's REST API that result in atomic database transactions on the running CockroachDB cluster. For details about the application code, see Implementation details.
Open the roach-data/roach-data-jdbc/src/main/resources/application.yml file and edit the datasource settings to connect to your running database cluster:
...
datasource:
url: jdbc:postgresql://localhost:{port}/roach_data?ssl=true&sslmode=require&sslrootcert={certs-dir}/ca.crt&sslkey={certs-dir}/client.{username}.key.pk8&sslcert={certs-dir}/client.{username}.crt
username: {username}
password: {password}
driver-class-name: org.postgresql.Driver
...
Where:
{port}is the port number.{certs-dir}is the full path to the certificates directory containing the authentication certificates that you created earlier.{username}and{password}specify the SQL username and password that you created earlier.
...
datasource:
url: jdbc:postgresql://{globalhost}:{port}/{cluster_name}.roach_data?sslmode=verify-full&sslrootcert={path to the CA certificate}/cc-ca.crt
username: {username}
password: {password}
driver-class-name: org.postgresql.Driver
...
Where:
{username}and{password}specify the SQL username and password that you created earlier.{globalhost}is the name of the CockroachDB Cloud free tier host (e.g.,free-tier.gcp-us-central1.cockroachlabs.cloud).{path to the CA certificate}is the path to thecc-ca.crtfile that you downloaded from the CockroachDB Cloud Console.{cluster_name}is the name of your cluster.
If you are using the connection string that you copied from the Connection info modal, your username, password, hostname, and cluster name will be pre-populated.
Open a terminal, and navigate to the roach-data-jdbc project subfolder:
$ cd <path>/roach-data/roach-data-jdbc
Use Maven to download the application dependencies and compile the code:
$ mvn clean install
From the roach-data-jdbc directory, run the application JAR file:
$ java -jar target/roach-data-jdbc.jar
The output should look like the following:
^__^
(oo)\_______
(__)\ )\/\ CockroachDB on Spring Data JDBC (v1.0.0.BUILD-SNAPSHOT)
||----w | powered by Spring Boot (v2.2.7.RELEASE)
|| ||
2020-06-17 14:56:54.507 INFO 43008 --- [ main] io.roach.data.jdbc.JdbcApplication : Starting JdbcApplication v1.0.0.BUILD-SNAPSHOT on MyComputer with PID 43008 (path/roach-data/roach-data-jdbc/target/roach-data-jdbc.jar started by user in path/roach-data/roach-data-jdbc)
2020-06-17 14:56:54.510 INFO 43008 --- [ main] io.roach.data.jdbc.JdbcApplication : No active profile set, falling back to default profiles: default
2020-06-17 14:56:55.387 INFO 43008 --- [ main] .s.d.r.c.RepositoryConfigurationDelegate : Bootstrapping Spring Data JDBC repositories in DEFAULT mode.
2020-06-17 14:56:55.452 INFO 43008 --- [ main] .s.d.r.c.RepositoryConfigurationDelegate : Finished Spring Data repository scanning in 59ms. Found 2 JDBC repository interfaces.
2020-06-17 14:56:56.581 INFO 43008 --- [ main] org.eclipse.jetty.util.log : Logging initialized @3378ms to org.eclipse.jetty.util.log.Slf4jLog
2020-06-17 14:56:56.657 INFO 43008 --- [ main] o.s.b.w.e.j.JettyServletWebServerFactory : Server initialized with port: 9090
2020-06-17 14:56:56.661 INFO 43008 --- [ main] org.eclipse.jetty.server.Server : jetty-9.4.28.v20200408; built: 2020-04-08T17:49:39.557Z; git: ab228fde9e55e9164c738d7fa121f8ac5acd51c9; jvm 11.0.7+10
2020-06-17 14:56:56.696 INFO 43008 --- [ main] o.e.j.s.h.ContextHandler.application : Initializing Spring embedded WebApplicationContext
2020-06-17 14:56:56.696 INFO 43008 --- [ main] o.s.web.context.ContextLoader : Root WebApplicationContext: initialization completed in 2088 ms
2020-06-17 14:56:57.170 INFO 43008 --- [ main] org.eclipse.jetty.server.session : DefaultSessionIdManager workerName=node0
2020-06-17 14:56:57.171 INFO 43008 --- [ main] org.eclipse.jetty.server.session : No SessionScavenger set, using defaults
2020-06-17 14:56:57.172 INFO 43008 --- [ main] org.eclipse.jetty.server.session : node0 Scavenging every 600000ms
2020-06-17 14:56:57.178 INFO 43008 --- [ main] o.e.jetty.server.handler.ContextHandler : Started o.s.b.w.e.j.JettyEmbeddedWebAppContext@deb3b60{application,/,[file:///private/var/folders/pg/r58v54857gq_1nqm_2tr6lg40000gn/T/jetty-docbase.3049902632643053896.8080/],AVAILABLE}
2020-06-17 14:56:57.179 INFO 43008 --- [ main] org.eclipse.jetty.server.Server : Started @3976ms
2020-06-17 14:56:58.126 INFO 43008 --- [ main] o.s.s.concurrent.ThreadPoolTaskExecutor : Initializing ExecutorService 'applicationTaskExecutor'
2020-06-17 14:56:58.369 INFO 43008 --- [ main] com.zaxxer.hikari.HikariDataSource : HikariPool-1 - Starting...
2020-06-17 14:56:58.695 INFO 43008 --- [ main] com.zaxxer.hikari.HikariDataSource : HikariPool-1 - Start completed.
2020-06-17 14:56:59.901 INFO 43008 --- [ main] liquibase.executor.jvm.JdbcExecutor : SELECT COUNT(*) FROM public.databasechangeloglock
2020-06-17 14:56:59.917 INFO 43008 --- [ main] liquibase.executor.jvm.JdbcExecutor : CREATE TABLE public.databasechangeloglock (ID INTEGER NOT NULL, LOCKED BOOLEAN NOT NULL, LOCKGRANTED TIMESTAMP WITHOUT TIME ZONE, LOCKEDBY VARCHAR(255), CONSTRAINT DATABASECHANGELOGLOCK_PKEY PRIMARY KEY (ID))
2020-06-17 14:56:59.930 INFO 43008 --- [ main] liquibase.executor.jvm.JdbcExecutor : SELECT COUNT(*) FROM public.databasechangeloglock
2020-06-17 14:56:59.950 INFO 43008 --- [ main] liquibase.executor.jvm.JdbcExecutor : DELETE FROM public.databasechangeloglock
2020-06-17 14:56:59.953 INFO 43008 --- [ main] liquibase.executor.jvm.JdbcExecutor : INSERT INTO public.databasechangeloglock (ID, LOCKED) VALUES (1, FALSE)
2020-06-17 14:56:59.959 INFO 43008 --- [ main] liquibase.executor.jvm.JdbcExecutor : SELECT LOCKED FROM public.databasechangeloglock WHERE ID=1
2020-06-17 14:56:59.969 INFO 43008 --- [ main] l.lockservice.StandardLockService : Successfully acquired change log lock
2020-06-17 14:57:01.367 INFO 43008 --- [ main] l.c.StandardChangeLogHistoryService : Creating database history table with name: public.databasechangelog
2020-06-17 14:57:01.369 INFO 43008 --- [ main] liquibase.executor.jvm.JdbcExecutor : CREATE TABLE public.databasechangelog (ID VARCHAR(255) NOT NULL, AUTHOR VARCHAR(255) NOT NULL, FILENAME VARCHAR(255) NOT NULL, DATEEXECUTED TIMESTAMP WITHOUT TIME ZONE NOT NULL, ORDEREXECUTED INTEGER NOT NULL, EXECTYPE VARCHAR(10) NOT NULL, MD5SUM VARCHAR(35), DESCRIPTION VARCHAR(255), COMMENTS VARCHAR(255), TAG VARCHAR(255), LIQUIBASE VARCHAR(20), CONTEXTS VARCHAR(255), LABELS VARCHAR(255), DEPLOYMENT_ID VARCHAR(10))
2020-06-17 14:57:01.380 INFO 43008 --- [ main] liquibase.executor.jvm.JdbcExecutor : SELECT COUNT(*) FROM public.databasechangelog
2020-06-17 14:57:01.396 INFO 43008 --- [ main] l.c.StandardChangeLogHistoryService : Reading from public.databasechangelog
2020-06-17 14:57:01.397 INFO 43008 --- [ main] liquibase.executor.jvm.JdbcExecutor : SELECT * FROM public.databasechangelog ORDER BY DATEEXECUTED ASC, ORDEREXECUTED ASC
2020-06-17 14:57:01.400 INFO 43008 --- [ main] liquibase.executor.jvm.JdbcExecutor : SELECT COUNT(*) FROM public.databasechangeloglock
2020-06-17 14:57:01.418 INFO 43008 --- [ main] liquibase.executor.jvm.JdbcExecutor : -- DROP TABLE IF EXISTS account cascade;
-- DROP TABLE IF EXISTS databasechangelog cascade;
-- DROP TABLE IF EXISTS databasechangeloglock cascade;
create table account
(
id int not null primary key default unique_rowid(),
balance numeric(19, 2) not null,
name varchar(128) not null,
type varchar(25) not null
)
2020-06-17 14:57:01.426 INFO 43008 --- [ main] liquibase.executor.jvm.JdbcExecutor : -- insert into account (id,balance,name,type) values
-- (1, 500.00,'Alice','asset'),
-- (2, 500.00,'Bob','expense'),
-- (3, 500.00,'Bobby Tables','asset'),
-- (4, 500.00,'Doris','expense');
2020-06-17 14:57:01.427 INFO 43008 --- [ main] liquibase.changelog.ChangeSet : SQL in file db/create.sql executed
2020-06-17 14:57:01.430 INFO 43008 --- [ main] liquibase.changelog.ChangeSet : ChangeSet classpath:db/changelog-master.xml::1::root ran successfully in 14ms
2020-06-17 14:57:01.430 INFO 43008 --- [ main] liquibase.executor.jvm.JdbcExecutor : SELECT MAX(ORDEREXECUTED) FROM public.databasechangelog
2020-06-17 14:57:01.441 INFO 43008 --- [ main] liquibase.executor.jvm.JdbcExecutor : INSERT INTO public.databasechangelog (ID, AUTHOR, FILENAME, DATEEXECUTED, ORDEREXECUTED, MD5SUM, DESCRIPTION, COMMENTS, EXECTYPE, CONTEXTS, LABELS, LIQUIBASE, DEPLOYMENT_ID) VALUES ('1', 'root', 'classpath:db/changelog-master.xml', NOW(), 1, '8:939a1a8c47676119a94d0173802d207e', 'sqlFile', '', 'EXECUTED', 'crdb', NULL, '3.8.9', '2420221402')
2020-06-17 14:57:01.450 INFO 43008 --- [ main] liquibase.executor.jvm.JdbcExecutor : INSERT INTO public.account (id, name, balance, type) VALUES ('1', 'Alice', 500.00, 'asset')
2020-06-17 14:57:01.459 INFO 43008 --- [ main] liquibase.changelog.ChangeSet : New row inserted into account
2020-06-17 14:57:01.460 INFO 43008 --- [ main] liquibase.executor.jvm.JdbcExecutor : INSERT INTO public.account (id, name, balance, type) VALUES ('2', 'Bob', 500.00, 'expense')
2020-06-17 14:57:01.462 INFO 43008 --- [ main] liquibase.changelog.ChangeSet : New row inserted into account
2020-06-17 14:57:01.462 INFO 43008 --- [ main] liquibase.executor.jvm.JdbcExecutor : INSERT INTO public.account (id, name, balance, type) VALUES ('3', 'Bobby Tables', 500.00, 'asset')
2020-06-17 14:57:01.464 INFO 43008 --- [ main] liquibase.changelog.ChangeSet : New row inserted into account
2020-06-17 14:57:01.465 INFO 43008 --- [ main] liquibase.executor.jvm.JdbcExecutor : INSERT INTO public.account (id, name, balance, type) VALUES ('4', 'Doris', 500.00, 'expense')
2020-06-17 14:57:01.467 INFO 43008 --- [ main] liquibase.changelog.ChangeSet : New row inserted into account
2020-06-17 14:57:01.469 INFO 43008 --- [ main] liquibase.changelog.ChangeSet : ChangeSet classpath:db/changelog-master.xml::2::root ran successfully in 19ms
2020-06-17 14:57:01.470 INFO 43008 --- [ main] liquibase.executor.jvm.JdbcExecutor : INSERT INTO public.databasechangelog (ID, AUTHOR, FILENAME, DATEEXECUTED, ORDEREXECUTED, MD5SUM, DESCRIPTION, COMMENTS, EXECTYPE, CONTEXTS, LABELS, LIQUIBASE, DEPLOYMENT_ID) VALUES ('2', 'root', 'classpath:db/changelog-master.xml', NOW(), 2, '8:c2945f2a445cf60b4b203e1a91d14a89', 'insert tableName=account; insert tableName=account; insert tableName=account; insert tableName=account', '', 'EXECUTED', 'crdb', NULL, '3.8.9', '2420221402')
2020-06-17 14:57:01.479 INFO 43008 --- [ main] l.lockservice.StandardLockService : Successfully released change log lock
2020-06-17 14:57:01.555 INFO 43008 --- [ main] o.s.b.a.e.web.EndpointLinksResolver : Exposing 8 endpoint(s) beneath base path '/actuator'
2020-06-17 14:57:01.610 INFO 43008 --- [ main] o.e.j.s.h.ContextHandler.application : Initializing Spring DispatcherServlet 'dispatcherServlet'
2020-06-17 14:57:01.610 INFO 43008 --- [ main] o.s.web.servlet.DispatcherServlet : Initializing Servlet 'dispatcherServlet'
2020-06-17 14:57:01.620 INFO 43008 --- [ main] o.s.web.servlet.DispatcherServlet : Completed initialization in 10 ms
2020-06-17 14:57:01.653 INFO 43008 --- [ main] o.e.jetty.server.AbstractConnector : Started ServerConnector@733c423e{HTTP/1.1, (http/1.1)}{0.0.0.0:9090}
2020-06-17 14:57:01.654 INFO 43008 --- [ main] o.s.b.web.embedded.jetty.JettyWebServer : Jetty started on port(s) 9090 (http/1.1) with context path '/'
2020-06-17 14:57:01.657 INFO 43008 --- [ main] io.roach.data.jdbc.JdbcApplication : Started JdbcApplication in 7.92 seconds (JVM running for 8.454)
2020-06-17 14:57:01.659 INFO 43008 --- [ main] io.roach.data.jdbc.JdbcApplication : Lets move some $$ around!
2020-06-17 14:57:03.552 INFO 43008 --- [ main] io.roach.data.jdbc.JdbcApplication : Worker finished - 7 remaining
2020-06-17 14:57:03.606 INFO 43008 --- [ main] io.roach.data.jdbc.JdbcApplication : Worker finished - 6 remaining
2020-06-17 14:57:03.606 INFO 43008 --- [ main] io.roach.data.jdbc.JdbcApplication : Worker finished - 5 remaining
2020-06-17 14:57:03.607 INFO 43008 --- [ main] io.roach.data.jdbc.JdbcApplication : Worker finished - 4 remaining
2020-06-17 14:57:03.608 INFO 43008 --- [ main] io.roach.data.jdbc.JdbcApplication : Worker finished - 3 remaining
2020-06-17 14:57:03.608 INFO 43008 --- [ main] io.roach.data.jdbc.JdbcApplication : Worker finished - 2 remaining
2020-06-17 14:57:03.608 INFO 43008 --- [ main] io.roach.data.jdbc.JdbcApplication : Worker finished - 1 remaining
2020-06-17 14:57:03.608 INFO 43008 --- [ main] io.roach.data.jdbc.JdbcApplication : Worker finished - 0 remaining
2020-06-17 14:57:03.608 INFO 43008 --- [ main] io.roach.data.jdbc.JdbcApplication : All client workers finished but server keeps running. Have a nice day!
As the output states, the application configures a database connection, starts a web servlet listening on the address http://localhost:9090/, initializes the account table and changelog tables with Liquibase, and then runs some test operations as requests to the application's REST API.
For more details about the application code, see Implementation details.
Query the database
Reads
The http://localhost:9090/account endpoint returns information about all accounts in the database. GET requests to these endpoints are executed on the database as SELECT statements.
The following curl command sends a GET request to the endpoint. The json_pp command formats the JSON response.
$ curl -X GET http://localhost:9090/account | json_pp
{
"_embedded" : {
"accounts" : [
{
"_links" : {
"self" : {
"href" : "http://localhost:9090/account/1"
}
},
"balance" : 500,
"name" : "Alice",
"type" : "asset"
},
{
"_links" : {
"self" : {
"href" : "http://localhost:9090/account/2"
}
},
"balance" : 500,
"name" : "Bob",
"type" : "expense"
},
{
"_links" : {
"self" : {
"href" : "http://localhost:9090/account/3"
}
},
"balance" : 500,
"name" : "Bobby Tables",
"type" : "asset"
},
{
"_links" : {
"self" : {
"href" : "http://localhost:9090/account/4"
}
},
"balance" : 500,
"name" : "Doris",
"type" : "expense"
}
]
},
"_links" : {
"self" : {
"href" : "http://localhost:9090/account?page=0&size=5"
}
},
"page" : {
"number" : 0,
"size" : 5,
"totalElements" : 4,
"totalPages" : 1
}
}
For a single account, specify the account number in the endpoint. For example, to see information about the accounts 1 and 2:
$ curl -X GET http://localhost:9090/account/1 | json_pp
{
"_links" : {
"self" : {
"href" : "http://localhost:9090/account/1"
}
},
"balance" : 500,
"name" : "Alice",
"type" : "asset"
}
$ curl -X GET http://localhost:9090/account/2 | json_pp
{
"_links" : {
"self" : {
"href" : "http://localhost:9090/account/2"
}
},
"balance" : 500,
"name" : "Bob",
"type" : "expense"
}
The http://localhost:9090/transfer endpoint performs transfers between accounts. POST requests to this endpoint are executed as writes (i.e., INSERTs and UPDATEs) to the database.
Writes
To make a transfer, send a POST request to the transfer endpoint, using the arguments specified in the "href" URL (i.e., http://localhost:9090/transfer%7B?fromId,toId,amount).
$ curl -X POST -d fromId=2 -d toId=1 -d amount=150 http://localhost:9090/transfer
You can use the accounts endpoint to verify that the transfer was successfully completed:
$ curl -X GET http://localhost:9090/account/1 | json_pp
{
"_links" : {
"self" : {
"href" : "http://localhost:9090/account/1"
}
},
"balance" : 350,
"name" : "Alice",
"type" : "asset"
}
$ curl -X GET http://localhost:9090/account/2 | json_pp
{
"_links" : {
"self" : {
"href" : "http://localhost:9090/account/2"
}
},
"balance" : 650,
"name" : "Bob",
"type" : "expense"
}
Monitor the application
http://localhost:9090/actuator is the base URL for a number of Spring Boot Actuator endpoints that let you monitor the activity and health of the application.
$ curl -X GET http://localhost:9090/actuator | json_pp
{
"_links" : {
"conditions" : {
"href" : "http://localhost:9090/actuator/conditions",
"templated" : false
},
"configprops" : {
"href" : "http://localhost:9090/actuator/configprops",
"templated" : false
},
"env" : {
"href" : "http://localhost:9090/actuator/env",
"templated" : false
},
"env-toMatch" : {
"href" : "http://localhost:9090/actuator/env/{toMatch}",
"templated" : true
},
"health" : {
"href" : "http://localhost:9090/actuator/health",
"templated" : false
},
"health-path" : {
"href" : "http://localhost:9090/actuator/health/{*path}",
"templated" : true
},
"info" : {
"href" : "http://localhost:9090/actuator/info",
"templated" : false
},
"liquibase" : {
"href" : "http://localhost:9090/actuator/liquibase",
"templated" : false
},
"metrics" : {
"href" : "http://localhost:9090/actuator/metrics",
"templated" : false
},
"metrics-requiredMetricName" : {
"href" : "http://localhost:9090/actuator/metrics/{requiredMetricName}",
"templated" : true
},
"self" : {
"href" : "http://localhost:9090/actuator",
"templated" : false
},
"threaddump" : {
"href" : "http://localhost:9090/actuator/threaddump",
"templated" : false
}
}
}
Each actuator endpoint shows specific metrics on the application. For example:
$ curl -X GET http://localhost:9090/actuator/health | json_pp
{
"components" : {
"db" : {
"details" : {
"database" : "PostgreSQL",
"result" : 1,
"validationQuery" : "SELECT 1"
},
"status" : "UP"
},
"diskSpace" : {
"details" : {
"free" : 125039620096,
"threshold" : 10485760,
"total" : 250685575168
},
"status" : "UP"
},
"ping" : {
"status" : "UP"
}
},
"status" : "UP"
}
For more information about actuator endpoints, see the Spring Boot Actuator Endpoint documentation.
Implementation details
This section guides you through the different components of the application project in detail.
Main application process
JdbcApplication.java defines the application's main process. It starts a Spring Boot web application, and then submits requests to the app's REST API that result in database transactions on the CockroachDB cluster.
Here are the contents of JdbcApplication.java:
package io.roach.data.jdbc;
import java.math.BigDecimal;
import java.util.ArrayDeque;
import java.util.Arrays;
import java.util.Deque;
import java.util.HashMap;
import java.util.LinkedList;
import java.util.Map;
import java.util.concurrent.ExecutionException;
import java.util.concurrent.Executors;
import java.util.concurrent.Future;
import java.util.concurrent.ScheduledExecutorService;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.springframework.boot.CommandLineRunner;
import org.springframework.boot.WebApplicationType;
import org.springframework.boot.autoconfigure.SpringBootApplication;
import org.springframework.boot.builder.SpringApplicationBuilder;
import org.springframework.context.annotation.EnableAspectJAutoProxy;
import org.springframework.core.Ordered;
import org.springframework.data.jdbc.repository.config.EnableJdbcRepositories;
import org.springframework.hateoas.Link;
import org.springframework.hateoas.config.EnableHypermediaSupport;
import org.springframework.transaction.annotation.EnableTransactionManagement;
import org.springframework.web.client.HttpStatusCodeException;
import org.springframework.web.client.RestTemplate;
/**
* Spring boot server application using spring-data-jdbc for data access.
*/
@EnableHypermediaSupport(type = EnableHypermediaSupport.HypermediaType.HAL)
@EnableJdbcRepositories
@EnableAspectJAutoProxy(proxyTargetClass = true)
@EnableTransactionManagement(order = Ordered.LOWEST_PRECEDENCE - 1) // Bump up one level to enable extra advisors
@SpringBootApplication
public class JdbcApplication implements CommandLineRunner {
protected static final Logger logger = LoggerFactory.getLogger(JdbcApplication.class);
public static void main(String[] args) {
new SpringApplicationBuilder(JdbcApplication.class)
.web(WebApplicationType.SERVLET)
.run(args);
}
@Override
public void run(String... args) {
final Link transferLink = Link.of("http://localhost:9090/transfer{?fromId,toId,amount}");
int concurrency = 1;
LinkedList<String> argsList = new LinkedList<>(Arrays.asList(args));
while (!argsList.isEmpty()) {
String arg = argsList.pop();
if (arg.startsWith("--concurrency=")) {
concurrency = Integer.parseInt(arg.split("=")[1]);
}
}
logger.info("Lets move some $$ around! (concurrency level {})", concurrency);
final ScheduledExecutorService executorService = Executors.newScheduledThreadPool(concurrency);
Deque<Future<Integer>> futures = new ArrayDeque<>();
for (int i = 0; i < concurrency; i++) {
Future<Integer> future = executorService.submit(() -> {
RestTemplate template = new RestTemplate();
int errors = 0;
for (int j = 0; j < 100; j++) {
int fromId = j % 4 + 1;
int toId = fromId % 4 + 1;
BigDecimal amount = new BigDecimal("10.00");
Map<String, Object> form = new HashMap<>();
form.put("fromId", fromId);
form.put("toId", toId);
form.put("amount", amount);
String uri = transferLink.expand(form).getHref();
logger.debug("({}) Transfer {} from {} to {}", uri, amount, fromId, toId);
try {
template.postForEntity(uri, null, String.class);
} catch (HttpStatusCodeException e) {
logger.warn(e.getResponseBodyAsString());
errors++;
}
}
return errors;
});
futures.add(future);
}
int totalErrors = 0;
while (!futures.isEmpty()) {
try {
int errors = futures.pop().get();
totalErrors += errors;
logger.info("Worker finished with {} errors - {} remaining", errors, futures.size());
} catch (InterruptedException e) {
Thread.currentThread().interrupt();
} catch (ExecutionException e) {
logger.warn("Worker failed", e.getCause());
}
}
logger.info("All client workers finished with {} errors and server keeps running. Have a nice day!",
totalErrors);
executorService.shutdownNow();
}
}
The annotations listed at the top of the JdbcApplication class definition declare some important configuration properties for the entire application:
@EnableHypermediaSupportenables hypermedia support for resource representation in the application. Currently, the only hypermedia format supported by Spring is HAL, and so thetype = EnableHypermediaSupport.HypermediaType.HAL. For details, see Hypermedia representation.@EnableJdbcRepositoriesenables the creation of Spring repositories for data access using Spring Data JDBC. For details, see Spring repositories.@EnableAspectJAutoProxyenables the use of@AspectJannotations for declaring aspects. For details, see Transaction management.@EnableTransactionManagementenables declarative transaction management in the application. For details, see Transaction management.Note that the
@EnableTransactionManagementannotation is passed anorderparameter, which indicates the ordering of advice evaluation when a common join point is reached. For details, see Ordering advice.@SpringBootApplicationis a standard configuration annotation used by Spring Boot applications. For details, see Using the @SpringBootApplication on the Spring Boot documentation site.
Schema management
To create and initialize the database schema, the application uses Liquibase.
Liquibase changelogs
Liquibase uses changelog files to manage database schema changes. Changelog files include a list of instructions, known as changesets, that are executed against the database in a specified order.
resources/db/changelog-master.xml defines the changelog for this application:
<?xml version="1.0" encoding="UTF-8"?>
<databaseChangeLog
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xmlns="http://www.liquibase.org/xml/ns/dbchangelog"
xsi:schemaLocation="http://www.liquibase.org/xml/ns/dbchangelog
http://www.liquibase.org/xml/ns/dbchangelog/dbchangelog-3.1.xsd">
<!-- Note: Context names match Spring profile names -->
<changeSet id="1" author="root" context="crdb">
<validCheckSum>ANY</validCheckSum>
<sqlFile path="db/create.sql"/>
</changeSet>
<changeSet id="2" author="root" context="crdb">
<insert tableName="account">
<column name="id">1</column>
<column name="name">Alice</column>
<column name="balance" valueNumeric="500.00"/>
<column name="type">asset</column>
</insert>
<insert tableName="account">
<column name="id">2</column>
<column name="name">Bob</column>
<column name="balance" valueNumeric="500.00"/>
<column name="type">expense</column>
</insert>
<insert tableName="account">
<column name="id">3</column>
<column name="name">Bobby Tables</column>
<column name="balance" valueNumeric="500.00"/>
<column name="type">asset</column>
</insert>
<insert tableName="account">
<column name="id">4</column>
<column name="name">Doris</column>
<column name="balance" valueNumeric="500.00"/>
<column name="type">expense</column>
</insert>
</changeSet>
</databaseChangeLog>
The first changeset uses the sqlFile tag, which tells Liquibase that an external .sql file contains some SQL statements to execute. The file specified by the changeset, resources/db/create.sql, creates the account table:
create table account
(
id int not null primary key default unique_rowid(),
balance numeric(19, 2) not null,
name varchar(128) not null,
type varchar(25) not null,
updated timestamptz not null default clock_timestamp()
);
-- insert into account (id,balance,name,type)
-- values
-- (1,100.50,'a','expense'),
-- (2,100.50,'b','expense'),
-- (3,100.50,'c','expense'),
-- (4,100.50,'d','expense'),
-- (5,100.50,'e','expense');
-- select * from account AS OF SYSTEM TIME '-5s';
The second changeset in the changelog uses the Liquibase XML syntax to specify a series of sequential INSERT statements that initialize the account table with some values.
When the application is started, all of the queries specified by the changesets are executed in the order specified by their changeset tag's id value. At application startup, Liquibase also creates a table called databasechangelog in the database where it performs changes. This table's rows log all completed changesets.
To see the completed changesets after starting the application, open a new terminal, start the built-in SQL shell, and query the databasechangelog table:
$ cockroach sql --certs-dir=certs
> SELECT * FROM roach_data.databasechangelog;
id | author | filename | dateexecuted | orderexecuted | exectype | md5sum | description | comments | tag | liquibase | contexts | labels | deployment_id
-----+--------+-----------------------------------+----------------------------------+---------------+----------+------------------------------------+--------------------------------------------------------------------------------------------------------+----------+------+-----------+----------+--------+----------------
1 | root | classpath:db/changelog-master.xml | 2020-06-17 14:57:01.431506+00:00 | 1 | EXECUTED | 8:939a1a8c47676119a94d0173802d207e | sqlFile | | NULL | 3.8.9 | crdb | NULL | 2420221402
2 | root | classpath:db/changelog-master.xml | 2020-06-17 14:57:01.470847+00:00 | 2 | EXECUTED | 8:c2945f2a445cf60b4b203e1a91d14a89 | insert tableName=account; insert tableName=account; insert tableName=account; insert tableName=account | | NULL | 3.8.9 | crdb | NULL | 2420221402
(2 rows)
Liquibase does not retry transactions automatically. If a changeset fails at startup, you might need to restart the application manually to complete the changeset.
Liquibase configuration
Typically, Liquibase properties are defined in a separate liquibase.properties file. In this application, the Spring properties file, application.yml, includes properties that enable and configure Liquibase:
...
liquibase:
change-log: classpath:db/changelog-master.xml
default-schema:
drop-first: false
contexts: crdb
enabled: true
...
The contexts property specifies a single Liquibase context (crdb). In order for a changeset to run, its context attribute must match a context set by this property. The context value is crdb in both of the changeset definitions in changelog-master.xml, so both changesets run at application startup.
For simplicity, application.yml only specifies properties for a single Spring profile, with a single set of Liquibase properties. If you want the changelog to include changesets that only run in specific environments (e.g., for debugging and development), you can create a new Spring profile in a separate properties file (e.g., application-dev.yml), and specify a different set of Liquibase properties for that profile. The profile set by the application configuration will automatically use the properties in that profile's properties file. For information about setting profiles, see the Spring documentation website.
Domain entities
Account.java defines the domain entity for the accounts table. This class is used throughout the application to represent a row of data in the accounts table.
Here are the contents of Account.java:
package io.roach.data.jdbc;
import java.math.BigDecimal;
import org.springframework.data.annotation.Id;
/**
* Domain entity mapped to the account table.
*/
public class Account {
@Id
private Long id;
private String name;
private AccountType type;
private BigDecimal balance;
public Long getId() {
return id;
}
public String getName() {
return name;
}
public AccountType getType() {
return type;
}
public BigDecimal getBalance() {
return balance;
}
}
Hypermedia representation
To represent database objects as HAL+JSON for the REST API, the application extends the Spring HATEOAS module's RepresentationModel class with AccountModel. Like the Account class, its attributes represent a row of data in the accounts table.
The contents of AccountModel.java:
package io.roach.data.jdbc;
import java.math.BigDecimal;
import org.springframework.hateoas.RepresentationModel;
import org.springframework.hateoas.server.core.Relation;
/**
* Account resource represented in HAL+JSON via REST API.
*/
@Relation(value = "account", collectionRelation = "accounts")
public class AccountModel extends RepresentationModel<AccountModel> {
private String name;
private AccountType type;
private BigDecimal balance;
public String getName() {
return name;
}
public void setName(String name) {
this.name = name;
}
public AccountType getType() {
return type;
}
public void setType(AccountType type) {
this.type = type;
}
public BigDecimal getBalance() {
return balance;
}
public void setBalance(BigDecimal balance) {
this.balance = balance;
}
}
We do not go into much detail about hypermedia representation in this tutorial. For more information, see the Spring HATEOAS Reference Documentation.
Spring repositories
To abstract the database layer, Spring applications use the Repository interface, or some subinterface of Repository. This interface maps to a database object, like a table, and its methods map to queries against that object, like a SELECT or an INSERT statement against a table.
AccountRepository.java defines the main repository for the accounts table:
package io.roach.data.jdbc;
import java.math.BigDecimal;
import org.springframework.data.jdbc.repository.query.Modifying;
import org.springframework.data.jdbc.repository.query.Query;
import org.springframework.data.repository.CrudRepository;
import org.springframework.data.repository.query.Param;
import org.springframework.stereotype.Repository;
import org.springframework.transaction.annotation.Transactional;
import static org.springframework.transaction.annotation.Propagation.MANDATORY;
/**
* The main account repository, notice there's no implementation needed since its auto-proxied by
* spring-data.
*/
@Repository
@Transactional(propagation = MANDATORY)
public interface AccountRepository extends CrudRepository<Account, Long>, PagedAccountRepository {
@Query(value = "SELECT balance FROM account WHERE id=:id FOR UPDATE")
BigDecimal getBalance(@Param("id") Long id);
@Modifying
@Query("UPDATE account SET balance = balance + :balance WHERE id=:id")
void updateBalance(@Param("id") Long id, @Param("balance") BigDecimal balance);
}
AccountRepository extends a subinterface of Repository that is provided by Spring for generic CRUD operations called CrudRepository. To support pagination queries, repositories in other Spring Data modules, like those in Spring Data JPA, usually extend a subinterface of CrudRepository, called PagingAndSortingRepository, that includes pagination and sorting methods. At the time this sample application was created, Spring Data JDBC did not support pagination. As a result, AccountRepository extends a custom repository, called PagedAccountRepository, to provide basic LIMIT/OFFSET pagination on queries against the accounts table. The AccountRepository methods use the @Query annotation strategy to define queries manually, as strings.
Note that, in addition to having the @Repository annotation, the AccountRepository interface has a @Transactional annotation. When transaction management is enabled in an application (i.e., with @EnableTransactionManagement), Spring automatically wraps all objects with the @Transactional annotation in a proxy that handles calls to the object. For more information, see Understanding the Spring Framework’s Declarative Transaction Implementation on Spring's documentation website.
@Transactional takes a number of parameters, including a propagation parameter that determines the transaction propagation behavior around an object (i.e., at what point in the stack a transaction starts and ends). This sample application follows the entity-control-boundary (ECB) pattern. As such, the REST service boundaries should determine where a transaction starts and ends rather than the query methods defined in the data access layer. To follow the ECB design pattern, propagation=MANDATORY for AccountRepository, which means that a transaction must already exist in order to call the AccountRepository query methods. In contrast, the @Transactional annotations on the Rest controller entities in the web layer have propagation=REQUIRES_NEW, meaning that a new transaction must be created for each REST request.
The aspects declared in TransactionHintsAspect.java and RetryableTransactionAspect.java further control how @Transactional-annotated components are handled. For more details on control flow and transaction management in the application, see Transaction management.
REST controller
There are several endpoints exposed by the application's web layer, some of which monitor the health of the application, and some that map to queries executed against the connected database. All of the endpoints served by the application are handled by the AccountController class, which is defined in AccountController.java:
package io.roach.data.jdbc;
import java.math.BigDecimal;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.dao.DataRetrievalFailureException;
import org.springframework.data.domain.PageRequest;
import org.springframework.data.domain.Pageable;
import org.springframework.data.domain.Sort;
import org.springframework.data.web.PageableDefault;
import org.springframework.data.web.PagedResourcesAssembler;
import org.springframework.hateoas.IanaLinkRelations;
import org.springframework.hateoas.Link;
import org.springframework.hateoas.PagedModel;
import org.springframework.hateoas.RepresentationModel;
import org.springframework.hateoas.UriTemplate;
import org.springframework.hateoas.server.RepresentationModelAssembler;
import org.springframework.http.HttpEntity;
import org.springframework.http.HttpStatus;
import org.springframework.http.ResponseEntity;
import org.springframework.transaction.annotation.Transactional;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.PathVariable;
import org.springframework.web.bind.annotation.PostMapping;
import org.springframework.web.bind.annotation.RequestParam;
import org.springframework.web.bind.annotation.RestController;
import org.springframework.web.servlet.support.ServletUriComponentsBuilder;
import static org.springframework.hateoas.server.mvc.WebMvcLinkBuilder.linkTo;
import static org.springframework.hateoas.server.mvc.WebMvcLinkBuilder.methodOn;
import static org.springframework.transaction.annotation.Propagation.REQUIRES_NEW;
/**
* Main remoting and transaction boundary in the form of a REST controller. The discipline
* when following the entity-control-boundary (ECB) pattern is that only service boundaries
* are allowed to start and end transactions. A service boundary can be a controller, business
* service facade or service activator (JMS/Kafka listener).
* <p>
* This is enforced by the REQUIRES_NEW propagation attribute of @Transactional annotated
* controller methods. Between the web container's HTTP listener and the transaction proxy,
* there's yet another transparent proxy in the form of a retry loop advice with exponential
* backoff. It takes care of retrying transactions that are aborted by transient SQL errors,
* rather than having these propagate all the way over the wire to the client / user agent.
*
* @see RetryableTransactionAspect
*/
@RestController
public class AccountController {
@Autowired
private AccountRepository accountRepository;
@Autowired
private PagedResourcesAssembler<Account> pagedResourcesAssembler;
/**
* Provides the service index resource representation which is only links
* for clients to follow.
*/
@GetMapping
public ResponseEntity<RepresentationModel> index() {
RepresentationModel index = new RepresentationModel();
// Type-safe way to generate URLs bound to controller methods
index.add(linkTo(methodOn(AccountController.class)
.listAccounts(PageRequest.of(0, 5)))
.withRel("accounts")); // Lets skip curies and affordances for now
index.add(Link.of(UriTemplate.of(linkTo(AccountController.class)
.toUriComponentsBuilder().path(
"/transfer/{?fromId,toId,amount}") // RFC-6570 template
.build().toUriString()),
"transfer"
).withTitle("Transfer funds"));
// Spring boot actuators for observability / monitoring
index.add(Link.of(
ServletUriComponentsBuilder
.fromCurrentContextPath()
.pathSegment("actuator")
.buildAndExpand()
.toUriString()
).withRel("actuator"));
return new ResponseEntity<>(index, HttpStatus.OK);
}
/**
* Provides a paged representation of accounts (sort order omitted).
*/
@GetMapping("/account")
@Transactional(propagation = REQUIRES_NEW)
public HttpEntity<PagedModel<AccountModel>> listAccounts(
@PageableDefault(size = 5, direction = Sort.Direction.ASC) Pageable page) {
return ResponseEntity
.ok(pagedResourcesAssembler.toModel(accountRepository.findAll(page), accountModelAssembler()));
}
/**
* Provides a point lookup of a given account.
*/
@GetMapping(value = "/account/{id}")
@Transactional(propagation = REQUIRES_NEW, readOnly = true) // Notice its marked read-only
public HttpEntity<AccountModel> getAccount(@PathVariable("id") Long accountId) {
return new ResponseEntity<>(accountModelAssembler().toModel(
accountRepository.findById(accountId)
.orElseThrow(() -> new DataRetrievalFailureException("No such account: " + accountId))),
HttpStatus.OK);
}
/**
* Main funds transfer method.
*/
@PostMapping(value = "/transfer")
@Transactional(propagation = REQUIRES_NEW)
public HttpEntity<BigDecimal> transfer(
@RequestParam("fromId") Long fromId,
@RequestParam("toId") Long toId,
@RequestParam("amount") BigDecimal amount
) {
if (amount.compareTo(BigDecimal.ZERO) < 0) {
throw new IllegalArgumentException("Negative amount");
}
if (fromId.equals(toId)) {
throw new IllegalArgumentException("From and to accounts must be different");
}
BigDecimal fromBalance = accountRepository.getBalance(fromId).add(amount.negate());
// Application level invariant check.
// Could be enhanced or replaced with a CHECK constraint like:
// ALTER TABLE account ADD CONSTRAINT check_account_positive_balance CHECK (balance >= 0)
if (fromBalance.compareTo(BigDecimal.ZERO) < 0) {
throw new NegativeBalanceException("Insufficient funds " + amount + " for account " + fromId);
}
accountRepository.updateBalance(fromId, amount.negate());
accountRepository.updateBalance(toId, amount);
return ResponseEntity.ok().build();
}
private RepresentationModelAssembler<Account, AccountModel> accountModelAssembler() {
return (entity) -> {
AccountModel model = new AccountModel();
model.setName(entity.getName());
model.setType(entity.getType());
model.setBalance(entity.getBalance());
model.add(linkTo(methodOn(AccountController.class)
.getAccount(entity.getId())
).withRel(IanaLinkRelations.SELF));
return model;
};
}
}
Annotated with @RestController, AccountController defines the primary web controller component of the application. The AccountController methods define the endpoints, routes, and business logic of REST services for account querying and money transferring. Its attributes include an instantiation of AccountRepository, called accountRepository, that establishes an interface to the accounts table through the data access layer.
As mentioned in the Spring repositories section, the application's transaction boundaries follow the entity-control-boundary (ECB) pattern, meaning that the web service boundaries of the application determine where a transaction starts and ends. To follow the ECB pattern, the @Transactional annotation on each of the HTTP entities (listAccounts(), getAccount(), and transfer()) has propagation=REQUIRES_NEW. This ensures that each time a REST request is made to an endpoint, a new transaction context is created. For details on how aspects handle control flow and transaction management in the application, see Transaction management.
Transaction management
When transaction management is enabled in an application, Spring automatically wraps all objects annotated with @Transactional in a proxy that handles calls to the object. By default, this proxy starts and closes transactions according to the configured transaction management behavior (e.g., the propagation level).
Using @AspectJ annotations, this sample application extends the default transaction proxy behavior with two other explicitly-defined aspects: TransactionHintsAspect and RetryableTransactionAspect. Methods of these aspects are declared as advice to be executed around method calls annotated with @Transactional.
For more information about transaction management in the app, see the following sections below:
Ordering advice
To determine the order of evaluation when multiple transaction advisors match the same pointcut (in this case, around @Transactional method calls), this application explicitly declares an order of precedence for calling advice.
At the top level of the application, in the main JdbcApplication.java file, the @EnableTransactionManagement annotation is passed an order parameter. This parameter sets the order on the primary transaction advisor to one level of precedence above the lowest level, Ordered.LOWEST_PRECEDENCE. This means that the advisor with the lowest level of precedence is evaluated after the primary transaction advisor (i.e., within the context of an open transaction).
For the two explicitly-defined aspects, TransactionHintsAspect and RetryableTransactionAspect, the @Order annotation is used. Like the order parameter on the @EnableTransactionManagement annotation, @Order takes a value that indicates the precedence of advice. The advisor with the lowest level of precedence is declared in TransactionHintsAspect, the aspect that defines the transaction attributes. RetryableTransactionAspect, the aspect that defines the transaction retry logic, defines the advisor with the highest level of precedence.
For more details about advice ordering, see Advice Ordering on the Spring documentation site.
Transaction attributes
The TransactionHintsAspect class, declared as an aspect with the @Aspect annotation, declares an advice method that defines the attributes of a transaction. The @Order annotation is passed the lowest level of precedence, Ordered.LOWEST_PRECEDENCE, indicating that this advisor must run after the main transaction advisor, within the context of a transaction. Here are the contents of TransactionHintsAspect.java:
package io.roach.data.jdbc;
import org.aspectj.lang.ProceedingJoinPoint;
import org.aspectj.lang.annotation.Around;
import org.aspectj.lang.annotation.Aspect;
import org.aspectj.lang.annotation.Pointcut;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.core.Ordered;
import org.springframework.core.annotation.Order;
import org.springframework.jdbc.core.JdbcTemplate;
import org.springframework.stereotype.Component;
import org.springframework.transaction.TransactionDefinition;
import org.springframework.transaction.annotation.Transactional;
import org.springframework.transaction.support.TransactionSynchronizationManager;
import org.springframework.util.Assert;
/**
* Aspect with an around advice that intercepts and sets transaction attributes.
* <p>
* This advice needs to runs in a transactional context, which is after the underlying
* transaction advisor.
*/
@Component
@Aspect
// After TX advisor
@Order(Ordered.LOWEST_PRECEDENCE)
public class TransactionHintsAspect {
protected final Logger logger = LoggerFactory.getLogger(getClass());
@Autowired
private JdbcTemplate jdbcTemplate;
private String applicationName = "roach-data";
@Pointcut("execution(* io.roach..*(..)) && @annotation(transactional)")
public void anyTransactionBoundaryOperation(Transactional transactional) {
}
@Around(value = "anyTransactionBoundaryOperation(transactional)",
argNames = "pjp,transactional")
public Object setTransactionAttributes(ProceedingJoinPoint pjp, Transactional transactional)
throws Throwable {
Assert.isTrue(TransactionSynchronizationManager.isActualTransactionActive(), "TX not active");
// https://www.cockroachlabs.com/docs/v19.2/set-vars.html
jdbcTemplate.update("SET application_name=?", applicationName);
if (transactional.timeout() != TransactionDefinition.TIMEOUT_DEFAULT) {
logger.info("Setting statement time {} for {}", transactional.timeout(),
pjp.getSignature().toShortString());
jdbcTemplate.update("SET statement_timeout=?", transactional.timeout() * 1000);
}
if (transactional.readOnly()) {
logger.info("Setting transaction read only for {}", pjp.getSignature().toShortString());
jdbcTemplate.execute("SET transaction_read_only=true");
}
return pjp.proceed();
}
}
The anyTransactionBoundaryOperation method is declared as a pointcut with the @Pointcut annotation. In Spring, pointcut declarations must include an expression to determine where join points occur in the application control flow. To help define these expressions, Spring supports a set of designators. The application uses two of them here: execution, which matches method execution joint points (i.e., defines a joint point when a specific method is executed, in this case, any method in the io.roach. namespace), and @annotation, which limits the matches to methods with a specific annotation, in this case @Transactional.
setTransactionAttributes sets the transaction attributes in the form of advice. Spring supports several different annotations to declare advice. The @Around annotation allows an advice method to work before and after the anyTransactionBoundaryOperation(transactional) join point. It also allows the advice method to call the next matching advisor with the ProceedingJoinPoint.proceed(); method.
On verifying that the transaction is active (using TransactionSynchronizationManager.isActualTransactionActive()), the advice sets some session variables using methods of the JdbcTemplate object declared at the top of the TransactionHintsAspect class definition. These session variables (application_name, statement_timeout, and transaction_read_only) set the application name for the query to "roach-data", the time allowed for the statement to execute before timing out to 1000 milliseconds (i.e., 1 second), and the transaction access mode as either READ ONLY or READ WRITE.
Transaction retries
Transactions may require retries if they experience deadlock or transaction contention that cannot be resolved without allowing serialization anomalies. To handle transactions that are aborted due to transient serialization errors, we highly recommend writing client-side transaction retry logic into applications written on CockroachDB.
Client-side retry handling is not necessary under READ COMMITTED isolation.
In this application, transaction retry logic is written into the methods of the RetryableTransactionAspect class. This class is declared an aspect with the @Aspect annotation. The @Order annotation on this aspect class is passed Ordered.LOWEST_PRECEDENCE-2, a level of precedence above the primary transaction advisor. This indicates that the transaction retry advisor must run outside the context of a transaction. Here are the contents of RetryableTransactionAspect.java:
package io.roach.data.jdbc;
import java.lang.reflect.UndeclaredThrowableException;
import java.util.concurrent.atomic.AtomicLong;
import org.aspectj.lang.ProceedingJoinPoint;
import org.aspectj.lang.annotation.Around;
import org.aspectj.lang.annotation.Aspect;
import org.aspectj.lang.annotation.Pointcut;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.springframework.core.Ordered;
import org.springframework.core.annotation.Order;
import org.springframework.dao.ConcurrencyFailureException;
import org.springframework.dao.TransientDataAccessException;
import org.springframework.stereotype.Component;
import org.springframework.transaction.TransactionSystemException;
import org.springframework.transaction.annotation.Transactional;
import org.springframework.transaction.support.TransactionSynchronizationManager;
import org.springframework.util.Assert;
/**
* Aspect with an around advice that intercepts and retries transient concurrency exceptions.
* Methods matching the pointcut expression (annotated with @Transactional) are retried a number
* of times with exponential backoff.
* <p>
* This advice needs to runs in a non-transactional context, which is before the underlying
* transaction advisor (@Order ensures that).
*/
@Component
@Aspect
// Before TX advisor
@Order(Ordered.LOWEST_PRECEDENCE - 2)
public class RetryableTransactionAspect {
protected final Logger logger = LoggerFactory.getLogger(getClass());
@Pointcut("execution(* io.roach..*(..)) && @annotation(transactional)")
public void anyTransactionBoundaryOperation(Transactional transactional) {
}
@Around(value = "anyTransactionBoundaryOperation(transactional)",
argNames = "pjp,transactional")
public Object retryableOperation(ProceedingJoinPoint pjp, Transactional transactional)
throws Throwable {
final int totalRetries = 30;
int numAttempts = 0;
AtomicLong backoffMillis = new AtomicLong(150);
Assert.isTrue(!TransactionSynchronizationManager.isActualTransactionActive(), "TX active");
do {
try {
numAttempts++;
return pjp.proceed();
} catch (TransientDataAccessException | TransactionSystemException ex) {
handleTransientException(ex, numAttempts, totalRetries, pjp, backoffMillis);
} catch (UndeclaredThrowableException ex) {
Throwable t = ex.getUndeclaredThrowable();
if (t instanceof TransientDataAccessException) {
handleTransientException(t, numAttempts, totalRetries, pjp, backoffMillis);
} else {
throw ex;
}
}
} while (numAttempts < totalRetries);
throw new ConcurrencyFailureException("Too many transient errors (" + numAttempts + ") for method ["
+ pjp.getSignature().toLongString() + "]. Giving up!");
}
private void handleTransientException(Throwable ex, int numAttempts, int totalAttempts,
ProceedingJoinPoint pjp, AtomicLong backoffMillis) {
if (logger.isWarnEnabled()) {
logger.warn("Transient data access exception (" + numAttempts + " of max " + totalAttempts + ") "
+ "detected (retry in " + backoffMillis + " ms) "
+ "in method '" + pjp.getSignature().getDeclaringTypeName() + "." + pjp.getSignature().getName()
+ "': " + ex.getMessage());
}
if (backoffMillis.get() >= 0) {
try {
Thread.sleep(backoffMillis.get());
} catch (InterruptedException e) {
Thread.currentThread().interrupt();
}
backoffMillis.set(Math.min((long) (backoffMillis.get() * 1.5), 1500));
}
}
}
The anyTransactionBoundaryOperation pointcut definition is identical to the one declared in TransactionHintsAspect. The execution designator matches all methods in the io.roach. namespace, and the @annotation designator limits the matches to methods with the @Transactional annotation.
retryableOperation handles the application retry logic in the form of advice. The @Around annotation allows the advice method to work before and after an anyTransactionBoundaryOperation(transactional) join point. It also allows the advice method to call the next matching advisor.
retryableOperation first verifies that there is no active transaction. It then increments the retry count and attempts to proceed to the next advice method with the ProceedingJoinPoint.proceed() method. If the underlying methods (i.e., the primary transaction advisor's methods and the annotated query methods) succeed, the transaction has been successfully committed to the database. The results are then returned and the application flow continues. If a failure in the underlying layers occurs due to a transient error (TransientDataAccessException or TransactionSystemException), then the transaction is retried. The time between each retry grows with each retry until the maximum number of retries is reached.
See also
Spring documentation:
- Spring Boot website
- Spring Framework Overview
- Spring Core documentation
- Data Access with JDBC
- Spring Web MVC
CockroachDB documentation:
- Learn CockroachDB SQL
- Client Connection Parameters
- CockroachDB Developer Guide
- Example Apps
- Transactions
Was this helpful?
