Implementing load balancing in Trino

For many of our use cases we are making use of Trino, for example connecting to Mongo, multi db connection when using with Open Source BI product Helical Insight etc. A Trino multi-node setup provides scalable, high-performance, and fault-tolerant data processing with efficient resource utilization and enhanced query parallelism.

In this blog we are going to cover how can you setup Trino on multiple load balancing mode so that the SQLquery execution can happen in parallel, thus helping in better performance as well as a fault tolerant server in case if one server goes down users requests will still be getting processed.

Setting Up a MultinodeTrino Cluster:

Setting up a multinodeTrinocluster involves configuring a coordinator node and one or more worker nodes. Here’s a step-by-step guide for setting up Trino on two servers: one as the coordinator and the other as the worker

1. Install trino on server1

Please refer to the following documentation link regarding how to install Trino

https://helicaltech.com/installing-trino-in-linux-server/

2. Configure Trino Coordinator:

Create the following configuration files in the etc directory of your Trino installation.

node.properties:

node.environment=production
node.id=ffffffff-ffff-ffff-ffff-ffffffffffff
node.data-dir=/var/trino/data

Note: The node.id should be a unique hexadecimal value

jvm.config:

-server
-Xmx4G
-XX:InitialRAMPercentage=80
-XX:MaxRAMPercentage=80
-XX:G1HeapRegionSize=32M
-XX:+ExplicitGCInvokesConcurrent
-XX:+ExitOnOutOfMemoryError
-XX:+HeapDumpOnOutOfMemoryError
-XX:-OmitStackTraceInFastThrow
-XX:ReservedCodeCacheSize=512M
-XX:PerMethodRecompilationCutoff=10000
-XX:PerBytecodeRecompilationCutoff=10000
-Djdk.attach.allowAttachSelf=true
-Djdk.nio.maxCachedBufferSize=2000000
-XX:+UnlockDiagnosticVMOptions
-XX:+UseAESCTRIntrinsics

config.properties:

coordinator=true
node-scheduler.include-coordinator=true
http-server.https.enabled=false
http-server.http.port=8080
query.max-memory=5GB
query.max-memory-per-node=1GB
discovery-server.enabled=true
discovery.uri=http://:8080
internal-communication.https.required=false
internal-communication.shared-secret=Testing@Helical2024

Note: The shared-secret should be the same in both the coordinator and worker configuration files.

log.properties:

io.trino=INFO

3. Start the Coordinatortrino server

Worker Configuration :

  1. Install trino on this server as well.
  2. Configure Trino Workers:

Create the following configuration files in the etc directory of your Trino installation on each worker node.

node.properties:

node.environment=production
node.id=9b43ef75-8cf0-4e4f-8aca-27eaf75fa83e
node.data-dir=/var/trino/data

Note: The node.id should be a unique hexadecimal value

jvm.config:

-server
-Xmx4G
-XX:InitialRAMPercentage=80
-XX:MaxRAMPercentage=80
-XX:G1HeapRegionSize=32M
-XX:+ExplicitGCInvokesConcurrent
-XX:+ExitOnOutOfMemoryError
-XX:+HeapDumpOnOutOfMemoryError
-XX:-OmitStackTraceInFastThrow
-XX:ReservedCodeCacheSize=512M
-XX:PerMethodRecompilationCutoff=10000
-XX:PerBytecodeRecompilationCutoff=10000
-Djdk.attach.allowAttachSelf=true
-Djdk.nio.maxCachedBufferSize=2000000
-XX:+UnlockDiagnosticVMOptions
-XX:+UseAESCTRIntrinsics

config.properties:

coordinator=false
http-server.http.port=8080
discovery.uri=http://:8080
internal-communication.shared-secret=Testing@Helical2024

Note: The shared-secret should be the same in both the coordinator and worker configuration files.

log.properties:

io.trino=INFO

3. Start the Workertrino

Thus we have configured 2 trino instances.

Verification and Management:

Access the Trino web UI at http://:8080 and ensure all workers are listed and active.

Note: The Trino port should be open between the two servers for communication.

Implementing load balancing in Trino

It should show 2 active workers

Implementing load balancing in Trino

Implementing load balancing in Trino

Please reach out to support@helicalinsight.com in case of any questions.

You May Also Like

Helical Insight’s self-service capabilities is one to reckon with. It allows you to simply drag and drop columns, add filters, apply aggregate functions if required, and create reports and dashboards on the fly. For advanced users, the self-service component has ability to add javascript, HTML, HTML5, CSS, CSS3 and AJAX. These customizations allow you to create dynamic reports and dashboards. You can also add new charts inside the self-service component, add new kind of aggregate functions and customize it using our APIs.
Helical Insight’s self-service capabilities is one to reckon with. It allows you to simply drag and drop columns, add filters, apply aggregate functions if required, and create reports and dashboards on the fly. For advanced users, the self-service component has ability to add javascript, HTML, HTML5, CSS, CSS3 and AJAX. These customizations allow you to create dynamic reports and dashboards. You can also add new charts inside the self-service component, add new kind of aggregate functions and customize it using our APIs.
Helical Insight, via simple browser based interface of Canned Reporting module, also allows to create pixel perfect printer friendly document kind of reports also like Invoice, P&L Statement, Balance sheet etc.
Helical Insight, via simple browser based interface of Canned Reporting module, also allows to create pixel perfect printer friendly document kind of reports also like Invoice, P&L Statement, Balance sheet etc.
If you have a product, built on any platform like Dot Net or Java or PHP or Ruby, you can easily embed Helical Insight within it using iFrames or webservices, for quick value add through instant visualization of data.
If you have a product, built on any platform like Dot Net or Java or PHP or Ruby, you can easily embed Helical Insight within it using iFrames or webservices, for quick value add through instant visualization of data.
Being a 100% browser-based BI tool, you can connect with your database and analyse across any location and device. There is no need to download or install heavy memory-consuming developer tools – All you need is a Browser application! We are battle-tested on most of the commonly used browsers.
Being a 100% browser-based BI tool, you can connect with your database and analyse across any location and device. There is no need to download or install heavy memory-consuming developer tools – All you need is a Browser application! We are battle-tested on most of the commonly used browsers.
We have organization level security where the Superadmin can create, delete and modify roles. Dashboards and reports can be added to that organization. This ensures multitenancy.
We have organization level security where the Superadmin can create, delete and modify roles. Dashboards and reports can be added to that organization. This ensures multitenancy.
We have organization level security where the Superadmin can create, delete and modify roles. Dashboards and reports can be added to that organization. This ensures multitenancy.
We have organization level security where the Superadmin can create, delete and modify roles. Dashboards and reports can be added to that organization. This ensures multitenancy.
A first-of-its-kind Open-Source BI framework, Helical Insight is completely API-driven. This allows you to add functionalities, including but not limited to adding a new exporting type, new datasource type, core functionality expansion, new charting in adhoc etc., at any place whenever you wish, using your own in-house developers.
A first-of-its-kind Open-Source BI framework, Helical Insight is completely API-driven. This allows you to add functionalities, including but not limited to adding a new exporting type, new datasource type, core functionality expansion, new charting in adhoc etc., at any place whenever you wish, using your own in-house developers.
It handles huge volumes of data effectively. Caching, Pagination, Load-Balancing and In-Memory not only provides you with amazing experience, but also and does not burden the database server more than required. Further effective use of computing power gives best performance and complex calculations even on the big data even with smaller machines for your personal use. Filtering, Sorting, Cube Analysis, Inter Panel Communication on the dashboards all at lightning speed. Thereby, making best open-source Business Intelligence solution in the market.
It handles huge volumes of data effectively. Caching, Pagination, Load-Balancing and In-Memory not only provides you with amazing experience, but also and does not burden the database server more than required. Further effective use of computing power gives best performance and complex calculations even on the big data even with smaller machines for your personal use. Filtering, Sorting, Cube Analysis, Inter Panel Communication on the dashboards all at lightning speed. Thereby, making best open-source Business Intelligence solution in the market.
With advance NLP algorithm, business users simply ask questions like, “show me sales of last quarter”, “average monthly sales of my products”. Let the application give the power to users without knowledge of query language or underlying data architecture
With advance NLP algorithm, business users simply ask questions like, “show me sales of last quarter”, “average monthly sales of my products”. Let the application give the power to users without knowledge of query language or underlying data architecture
Our application is compatible with almost all databases, be it RDBMS, or columnar database, or even flat files like spreadsheets or csv files. You can even connect to your own custom database via JDBC connection. Further, our database connection can be switched dynamically based on logged in users or its organization or other parameters. So, all your clients can use the same reports and dashboards without worrying about any data security breech.
Our application is compatible with almost all databases, be it RDBMS, or columnar database, or even flat files like spreadsheets or csv files. You can even connect to your own custom database via JDBC connection. Further, our database connection can be switched dynamically based on logged in users or its organization or other parameters. So, all your clients can use the same reports and dashboards without worrying about any data security breech.
Our application can be installed on an in-house server where you have full control of your data and its security. Or on cloud where it is accessible to larger audience without overheads and maintenance of the servers. One solution that works for all.
Our application can be installed on an in-house server where you have full control of your data and its security. Or on cloud where it is accessible to larger audience without overheads and maintenance of the servers. One solution that works for all.
Different companies have different business processes that the existing BI tools do not encompass. Helical Insight permits you to design your own workflows and specify what functional module of BI gets triggered
Different companies have different business processes that the existing BI tools do not encompass. Helical Insight permits you to design your own workflows and specify what functional module of BI gets triggered