Dawiso Integration Runtime (DIR)

Dawiso Integration Runtime (DIR) is the on-premises alternative for cloud-level scanning of metadata into Dawiso.

DIR is suitable in the following scenarios:

• Your metadata source (i.e. database engine, API, etc.) is in your infrastructure behind a firewall and access to it cannot be opened for the public internet.
• You cannot share access credentials to your metadata sources with Dawiso because of security policies or any other concerns. See Connection types for more details.
• To optimize performance, a large number of metadata assets should be scanned locally first instead of scanning it from a Dawiso cloud instance.

Understanding the DIR workflow

1. DIR runs in your infrastructure and connects to your cloud or on-premise data sources with the connection properties and credentials.
   • Properties and credentials were specified in the DIR configuration JSON, either as literal values or as environment variable placeholders.
2. Metadata is ingested from the source into JSON files. Files are stored in a local path and compressed with a standard gzip algorithm.
   • You can review the content and see what exactly was extracted from the metadata source and it will be sent to Dawiso if needed.
3. Once metadata extraction is completed, all gzipped files will be sent to your Dawiso cloud instance Integration API using standard HTTPS protocol (port 443).
4. Dawiso will process incoming files and import metadata in the target Space & Application metadata repository.

Requirements to run DIR

To successfully run Dawiso Integration Runtime (DIR), your environment must meet several system and network requirements. These ensure reliable metadata ingestion and smooth performance during scanning operations. The following sections outline the necessary prerequisites for installation and execution of DIR.

1. .NET 10 Runtime

DIR is a .NET 10 application. It can be run on both Windows and Linux systems, provided that the .NET 10 runtime is installed.

Alternatively, run DIR as a Docker container — the image bundles the runtime, so no local .NET installation is needed. See Run DIR in Docker.

2. Hardware Requirements

The following hardware sizing is recommended based on the size of your metadata environment:

Standard environments	Large environments
2 vCPUs, 4 GB RAM Sufficient for most common metadata scanning scenarios.	4 vCPUs, 8 GB RAM For datasets exceeding 5 million metadata assets, higher hardware specifications are recommended to ensure better performance and processing capacity.

3. Local Storage

DIR requires local storage to temporarily store metadata extraction files.

• Typical usage: A few gigabytes of space are sufficient.
• For large environments or longer retention periods: Provision up to 500 GB of storage to ensure stability and future scalability.

DIR deletes files from previous runs on its own for a data source that keeps running successfully, which keeps storage use from growing indefinitely under normal conditions. See Clean up previous runs for what it keeps, and the cases where cleanup doesn’t happen.

4. Network Access

DIR must be able to access the following network endpoints:

• Metadata sources:
  • A direct network connection to each metadata source is required.
  • Authentication methods and supported connection types vary per source (see individual connector documentation).
• Dawiso Cloud: An HTTPS (port 443) connection is required to communicate with the Dawiso Cloud Integration API.

Download DIR to your local device

Download DIR from this link. Select DIR version based on your machine.

DIR file	When to use
integration-runtime-linux.zip	For use on Linux systems.
integration-runtime-windows.zip	Select this version if you have .NET 10.0 on your local machine. If necessary, you might receive a prompt with links and instructions on how to download .NET.
integration-runtime-windows-self-contained.zip	Select this all-included version if you do not have admin permissions to install .NET on your local machine.

Run DIR in Docker

To run DIR as a container — for example, on a Linux server or as part of a scheduled pipeline — use the official Docker image instead of the zip download. The image bundles the .NET runtime, so nothing needs to be installed on the host.

Pull the image

The DIR container image is hosted in Dawiso’s container registry. Dawiso provides you with read-only credentials to pull the image:

docker login dawisopublic.azurecr.io
docker pull dawisopublic.azurecr.io/di-runtime:latest

Run the container

1. Place your data source configuration JSON on the host and create a directory to serve as the working folder.

2. Run the container, mounting the configuration file and the working folder:

   docker run --rm \
     -v "/path/to/dir-config.json:/config/dir-config.json" \
     -v "/path/to/working-folder:/App/WorkingFolder" \
     dawisopublic.azurecr.io/di-runtime:latest --config /config/dir-config.json

   Tip

   On Windows hosts, use Windows paths on the left-hand side of each -v mount, for example: -v "C:\DIR\dir-config.json:/config/dir-config.json"

3. Wait for the ingestion to finish, then check the console output or the mounted working folder to confirm the run was successful.

For regular scans, use the built-in Cron setting or an external scheduler. For more information, see Schedule ingestion runs.

For a step-by-step guide including how to obtain the data source configuration JSON, see Private connection data ingestion (DIR).

Schedule ingestion runs

By default, DIR runs a single ingestion and exits. To ingest on a schedule, use the built-in Cron setting or an external scheduler.

Built-in scheduling with Cron

Add the Cron property to the General section of your configuration JSON:

"General": {
    "WorkingFolder": "WorkingFolder",
    "ExtractOnly": false,
    "Cron": "0 2 * * * ! Europe/Prague"
}

When Cron is set, DIR stays running and starts an ingestion on each occurrence.

Property	Description	Values	Mandatory
Cron	Standard five-field cron expression, optionally suffixed with ! <timezone> (IANA name). Without the suffix, the schedule runs in UTC.	String	No
RunImmediately	When true together with Cron, runs an ingestion immediately at startup and then continues on the schedule. Defaults to false.	true / false	No

If an occurrence falls while an ingestion is still running, the next run starts after the current one finishes, with a minimum gap of 5 minutes between runs.

When running in Docker with Cron set, start the container detached instead of using --rm, so it keeps running between occurrences:

docker run -d --name dawiso-dir --restart unless-stopped \
  -v "/path/to/dir-config.json:/config/dir-config.json" \
  -v "/path/to/working-folder:/App/WorkingFolder" \
  dawisopublic.azurecr.io/di-runtime:latest --config /config/dir-config.json

External scheduler

Alternatively, leave Cron unset and schedule the one-shot run with cron (Linux), Task Scheduler (Windows), or your orchestrator. See Private connection data ingestion (DIR) for a Task Scheduler example.

Upgrade DIR

Upgrading DIR is necessary when changes have been made to ingestion that affect your environment, for example, if updates were made to the connectors you are using or if new connectors were developed that you want to take advantage of.

To upgrade DIR, follow these steps:

1. Download the newest version of DIR from one of the links above.
2. In the existing DIR folder, make sure the following are preserved:
   • WorkingFolder
   • All DawisoIntegrationRuntime files (= DIR configuration files)
3. Delete all other DIR files.
4. Extract DIR to the original folder with the WorkingFolder and DIR configuration files.

Upgrade in Docker

When you run DIR in Docker, pull the newer image instead:

docker pull dawisopublic.azurecr.io/di-runtime:latest

The configuration JSON and the working folder are mounted from the host, so they are preserved automatically.

Clean up previous runs

From version 2026.3, DIR deletes files from previous runs in the WorkingFolder automatically. Cleanup runs at the end of every successful run. A failed run deletes nothing, so its log and archive stay available for troubleshooting.

DIR cleans up three kinds of files:

Files	Setting	Default	What DIR keeps
Run logs (Logs/Ingestion#*.log)	LogRetentionDays	30	Logs younger than the set number of days, and always the 5 newest logs
Run archives (*.zip)	KeepRunArchives	5	The newest archives, up to the set number. An archive that was not uploaded to Dawiso is never deleted, so you can still upload it again
Backups created by a run with --restart (*.backup-* folders)	BackupRetentionDays	30	Backups younger than the set number of days, and always the newest backup

When ExtractOnly is set to true, the archive is the result of the run and is never uploaded. DIR then keeps only the set number of newest archives, whether they were uploaded or not.

To change the defaults, add a Retention block to the General section of the DIR configuration JSON:

{
    "General": {
        "WorkingFolder": "...",
        "Retention": {
            "LogRetentionDays": 30,
            "KeepRunArchives": 5,
            "BackupRetentionDays": 30
        }
    }
}

DIR writes every deleted file and the space it freed to the log. If a file cannot be deleted, DIR writes a warning and the run still succeeds.

Ingest data using DIR

For step-by-step guides on how to ingest data using DIR, see:

• Ingest data with a private connection using DIR
• Ingest data with Active Directory authentication