Collect logs for CAST Support

Identify the log CAST Support is asking for and retrieve it from the CAST Imaging UI or from disk on Microsoft Windows and Linux

Overview

When you raise a case, CAST Support often asks for one or more logs. This page lists the names Support uses for each log, then explains how to retrieve it - from the CAST Imaging UI where possible, and from disk on Microsoft Windows and Linux via Docker/Podman where you have access to the servers.

Which log does Support mean?

Find the name Support used in the left column, then go to the section listed in the right column. Where the name is ambiguous, ask Support which one they mean.

Name used by Support What it is How to get it
Analysis log, job log, application log, AIP log, CAST-MS log, Core log The step-by-step log of a job run on an application: source code delivery, fast scan, analysis, snapshot, publication to the viewer Analysis logs
A step name or step ID, for example “generate views (process_imaging) logs” The log of one step of a job Logs of a specific step
Debug log, analysis log in debug mode An analysis log produced with debug messages turned on Analysis logs in debug mode
Node log, analysis-node log, analysis-node.log The service log of the analysis-node component, not the log of an analysis Service logs
Console log, console service log, console.log, aip-gateway.log The log of the console service in imaging-services Service logs
Browser console, developer console, F12 log Errors displayed by your web browser, not by CAST Imaging Browser logs
HAR, HAR file, network trace A recording of the requests your browser exchanged with CAST Imaging Browser logs
Gateway log The gateway service in imaging-services: every request to CAST Imaging goes through it Service logs
Auth log, auth-service log The auth-service service in imaging-services Service logs
SSO log, Keycloak log The sso service in imaging-services (Keycloak), which handles login, LDAP and SAML Service logs
Control Panel log, CP log, admin center log The control-panel service in imaging-services, which registers the other services Service logs
Viewer log, server log, imaging-service log, access log The server service in imaging-viewer, which serves the viewer UI Service logs
ETL log, imaging-etl.log The etl service in imaging-viewer, which loads analysis results into Neo4j at the end of an analysis Service logs
API log, Imaging API log The imaging_apis service in imaging-viewer Service logs
AI Service log, AI Manager log, agents log, GraphRAG log The open_ai_manager service in imaging-viewer, which serves the AI features Service logs
Neo4j log, neo4j.log, debug.log, query.log The graph database of imaging-viewer Service logs
Dashboards log, restapi.log The dashboards component Service logs
Container log, stdout, docker logs What a container writes to its standard output, on Linux only Container logs
Install log, installer log, update log The output of the installer or the update Installation and update logs
PostgreSQL log The PostgreSQL instance that holds the analysis schemas Other logs
Extend proxy log, Extend local server log CAST Extend local server Other logs
MCP server log CAST Imaging MCP Server or Gatekeeper MCP Server Other logs
Application backup, delivery folder, schema backup Not a log: a copy of the application’s delivery folder and analysis schemas, which Support uses to reproduce a problem Delivery folder and schema backup

Before you send logs

Include the following with the logs so Support can find the relevant lines:

  • The CAST Imaging release, for example 3.6.9-funcrel.
  • The date, time and time zone at which the problem occurred, and the name of the application if one is involved.
  • The name of the node that ran the job, in a deployment with several nodes.

Choose a date range that covers the problem. Logs contain the paths and names of your source files, and can contain host names and user names: review them against your own policy before you send them.

Deployments with several machines

In a distributed deployment, the CAST Imaging components are spread over several machines, for example imaging-services and dashboards on one machine, imaging-viewer on a second, and one or more analysis-node instances on others. Each component writes its logs on the machine where it runs, so a log you retrieve from disk is only on one of them.

The UI does not have this constraint: the application’s Logs page and the Logs tab of the Services settings page retrieve logs from every machine, so you can download everything from your browser without knowing where each component runs. Use them where you can.

When you retrieve logs from disk, use this table to find the machine:

Logs Machine that holds them
Analysis logs, job-logs copy None in particular: job-logs is in the shared common-data folder, which every node and the imaging-services machine can read
Analysis logs, Logs folder, including the logs of the tools such as CAST-MS The node that ran the job only. The Logs page shows the node of each job as <host>:<port>
Node log Every node, one log each. Send the log of the node that ran the job, or the logs of all nodes when the problem is not tied to a job
imaging-services logs: Control Panel, Gateway, Auth Service, Console, SSO The machine where imaging-services is installed
imaging-viewer logs: Viewer, ETL, Imaging API, AI Service, Neo4j The machine where imaging-viewer is installed
Dashboards log The machine where dashboards is installed
Container logs The machine that runs the container
Installation and update logs The machine on which the installer or the update was run

Analysis logs

Analysis logs record each step of a job run on an application: one log per step, plus the logs of the tools the step runs, such as CAST-MS, DMT or the profiler.

From the application’s Logs page

You need access to the application.

  1. Open the application, then select Logs in the left menu. The page lists the jobs run on the application, most recent first, each with its start time, duration and the node that ran it. See Logs page for a description of the page.
  2. Find the job in question. A deep analysis is listed as Configure - Analyze - Upload, a fast scan as Fast Scan, and the dashboards consolidation that follows a deep analysis as a separate job named Upload.
  3. Download what Support asked for:
    • The whole job: select the download icon at the end of the job’s row. You receive one ZIP file, for example 20261111-132042_deep_analysis_logs.zip, with one folder per step named with the step ID. When in doubt, send this, or, if you are an administrator, the more complete download described in From the application details, for administrators.
    • One step: expand the job, then select the download icon at the end of the step’s row. Steps are listed with the names given in Logs of a specific step.
    • One log of a step, for example the log of one analyzer: select the eye icon of the step to open it, select the log in the left panel, then select the download icon of the LOGS tab.

In this example, the deep analysis job is expanded to show its steps, with the dashboards consolidation (Upload) above it and the Fast Scan below:

The eye icon of a step opens it, with the logs of the tools it ran in the left panel. The download icon of the LOGS tab downloads the log selected on the left:

The Job Monitor uses the same job names as the Logs page, which helps you find the right job when several have run.

From the application details, for administrators

You need the Admin profile or a profile with the Administrator role. This download is the most complete record of a job, and does not require access to the application:

  1. Select the application in the list of applications, then open its details from the Actions menu - see Managing Application details.
  2. Expand the CAST Support section. It lists the jobs run on the application, most recent first, with the version, the node that ran each job, its status and its date.
  3. Select the download icon at the end of the job’s row.

You receive one ZIP file named after the job and its start time: <date>-<time>_fast_scan_original_logs.zip, <date>-<time>_deep_analysis_original_logs.zip or <date>-<time>_run_measure_task_original_logs.zip. The Logs page download of the same job is named ..._logs.zip instead.

Both ZIP files have one folder per step, named with the step ID, and the same logs in each step folder. They differ in form:

Logs page download CAST Support download
Log files Converted to plain .log files, numbered in the order they ran, for example 18-jeeanalyzer.log The original files, with their original names and formats, for example Carl_MA_JEE.log or DMT-CreateVersionForServer-<date>.castlog2
castms-logs folder No The working logs of CAST-MS, including the debug logs and the per-analyzer reports, for a deep analysis or an Upload job
dmt-delivery folder No The logs written in the delivery folder when the source code was packaged, for a deep analysis
log-file-list.txt No The full path on the server of every file in the ZIP

The section names jobs differently from the Logs page:

CAST Support section Logs page
Discover Fast Scan
Configure - Analyze - Publish Configure - Analyze - Upload
Publish Upload

From disk

Each job has an ID, and its logs are stored in two places:

  • In the job-logs folder of the shared common-data folder - one subfolder per job, then one subfolder per step, named with the step ID. This is the copy the Logs page displays, and the easiest one to send: compress the job’s folder.
  • In the Logs folder of the node that ran the job - one subfolder per application, holding one subfolder per job with one file per step, named 0-<step>.log. The logs of the tools run by each step, such as CAST-MS-*.log, DMT-*.castlog2 and profiler-*.log, are at the application level.
Location Microsoft Windows Linux via Docker/Podman
job-logs <common-data>\job-logs\<job ID>\<step ID>\ /opt/cast/shared/common-data/job-logs/<job ID>/<step ID>/
Logs on the node %PROGRAMDATA%\CAST\CAST\Logs\<application>\<job ID>\ /opt/cast/core/CAST/Logs/<application>/<job ID>/ - inside the analysis-node container, /usr/share/CAST/CAST/Logs

On Microsoft Windows, <common-data> is the folder set with SHARED_FOLDER when CAST Imaging was installed, by default c:\cast-node\common-data. To find it, run the following in PowerShell:

Select-String -Path "$env:ProgramData\CAST\CAST\.install\all-currenSettings.txt" -Pattern 'SHARED_FOLDER'

For example, a fast scan of the application JSP produced:

job-logs/29bfa847-810e-438d-a4c3-c8f7560109b8/
  code_scanner/
  create_delivery_folder/
  prepare_version/
  unzip_source/

Logs/JSP/
  29bfa847-810e-438d-a4c3-c8f7560109b8/
    0-code-scanner.log
    0-create-delivery-folder.log
    0-prepare-version.log
    0-unzip-source.log
  CAST-MS-createEmptyApplication-20261109-135802.log
  DMT-CreateVersionForServer-20261109-135816.castlog2
  profiler-20261109-135835.log

To find the ID of a job, use the most recent folder in job-logs, or search the node log for the date of the job: each job starts with a line such as Starting 'fast_scan' job execution '<job ID>'. In a deployment with several nodes, the job may have run on any of them, but job-logs is on the shared folder that all nodes use - see Deployments with several machines.

Logs of a specific step

Support often names the step whose log they need, using either the name shown on the Logs page or the step ID. The step ID is also the name of the step’s folder in job-logs, and the step’s main log is 0-<step ID with dashes>.log. For example, “generate views (process_imaging) logs” is the log of the Generate views step: download it from the Logs page, or send the folder job-logs/<job ID>/process_imaging/.

Each step also holds the logs of the tools it ran, numbered in the order they ran: CAST-MS, ServMan, DMT, the profiler, or one log per analyzer and extension for Run analysis. On disk, send the whole step folder, not only the 0- file.

Fast Scan job

Shown on the Logs page Step ID What the step does
Create delivery folder create_delivery_folder Creates the application
Unzipping source unzip_source Extracts the uploaded source code
Initialize fast scan prepare_version Creates the version
Content discovery code_scanner Scans and classifies the source files

Configure - Analyze - Upload job

The deep analysis:

Shown on the Logs page Step ID What the step does
Exclude files exclude_files Excludes the files selected for exclusion
Prepare environment prepare_environment Creates the application schemas and imports preferences
Install extensions install_extensions Installs the extensions in the application schemas
Prepare analysis config prepare_analysis_config Packages, delivers and accepts the version
Inject analysis options inject_analysis_options Applies the analysis settings
Run analysis analyze Runs the analyzers and extensions
Finalizing analysis restore_files Restores the files excluded before the analysis
Compute the structural flaws snapshot Computes the snapshot
Generate snapshot indicators snapshot_indicator Computes the snapshot indicators
Prepare publication to Management Dashboard upload_application Sends the snapshot to the Measurement Service, which then starts the Upload job. When results are missing from the dashboards, send this step with the Upload job that followed it and the dashboards log
Generate views process_imaging Sends the results to the etl service of imaging-viewer, which loads them into Neo4j for the viewer. The step log includes the ETL’s output for the import: when the cause is not there, also send the ETL and Neo4j logs for the same date range

Upload job

The consolidation for the dashboards, started by the Prepare publication to Management Dashboard step of the deep analysis. Its job type is run_measure_task:

Shown on the Logs page Step ID What the step does
Publish to Management Dashboard run_measure_sync Loads the snapshot into the measurement schema and consolidates it for the dashboards, with its own aad-consolidation-*.log. Send it with the Prepare publication to Management Dashboard step and the dashboards log

The job folder also holds log-alerts.json, which lists the alerts raised during the job, such as missing JAR files. Support may ask for it with the step logs.

Analysis logs in debug mode

Where Support asks for a debug log, turn on debug messages, run the job again, then retrieve the analysis log as above:

  1. In the application settings, turn on Include debug messages - see Using the debug options.
  2. Run the job again.
  3. Retrieve the analysis log.

The option turns itself off after the next analysis. Turn on Show SQL or Activate AMT Memory Profiling only when Support asks for them: both slow the analysis down.

Service logs

Service logs record the activity of the CAST Imaging components themselves: login, requests, communication between services, loading results into the viewer. They are kept per service, not per application.

From the UI

Available in ≥ 3.6.9-funcrel

You need the Admin profile or a profile with the Administrator role.

  1. Select Settings, then Services, then the Logs tab.
  2. Choose a date range of up to 31 days that covers the problem.
  3. Select the services Support asked for, or select the checkbox in the Service header to select them all.
  4. Select DOWNLOAD. Several logs arrive as a single ZIP file with one folder per service.

See Logs tab for details. In 3.6.8-funcrel, each service has its own Service logs panel on the Services tab instead - see Download service logs in 3.6.8-funcrel. In earlier releases, only the node log can be downloaded from the UI: retrieve the others from disk.

The Logs tab names each service as follows:

Component Name on the Logs tab Support may call it
analysis-nodes One entry per node Node log
imaging-services Control Panel, Gateway, Auth Service, Console, SSO Control Panel, Gateway, Auth, Console, SSO or Keycloak log
imaging-viewer Imaging Viewer, ETL, Imaging API, AI Service, Neo4j Viewer or server, ETL, API, AI Manager, Neo4j log
dashboards The dashboards service Dashboards log

The Logs tab downloads service logs only. A node entry, for example, arrives as <host>-<port>_node-logs_<from>_<to>.zip and contains the node’s service log, analysis-node.log, but no analysis logs. Retrieve analysis logs from the application’s Logs page or from disk - see Analysis logs.

From disk on Microsoft Windows

Service Default location Files
Node %PROGRAMDATA%\CAST\Imaging\CAST-Imaging-Analysis-Node\logs analysis-node.log
SSO %PROGRAMDATA%\CAST\Imaging\CAST-Imaging-SSO\logs sso.log for the current day and one .gz file per previous day, plus CAST-Imaging-SSO-log.txt, written by the Windows service that starts SSO
Console %PROGRAMDATA%\CAST\Imaging\CAST-Imaging-Console\logs aip-gateway.log - despite its name, this is the log of the Console service
Gateway %PROGRAMDATA%\CAST\Imaging\CAST-Imaging-Gateway gateway.log
Auth Service %PROGRAMDATA%\CAST\Imaging\CAST-Imaging-Auth-Service auth-service.log
Control Panel %PROGRAMDATA%\CAST\Imaging\CAST-Imaging-Control-Panel control-panel.log
Viewer %PROGRAMFILES%\CAST\Imaging\CAST-Imaging-Viewer\logs imaging-service.log, nginx_access.log, nginx_error.log, error.log
ETL %PROGRAMFILES%\CAST\Imaging\CAST-Imaging-Viewer\logs imaging-etl.log, and one import log per application, such as <application>_local-default-<number>.log
Imaging API %PROGRAMFILES%\CAST\Imaging\CAST-Imaging-Viewer\logs imaging-api.log
AI Service %PROGRAMFILES%\CAST\Imaging\CAST-Imaging-Viewer\logs ai_service.log, imaging-agents.log, agents_server.log, graphrag_server.log, imaging_graphrag.log
Neo4j %PROGRAMFILES%\CAST\Imaging\CAST-Imaging-Viewer\logs neo4j.log, debug.log, query.log, security.log, neo4j-service.<date>.log
Dashboards %PROGRAMDATA%\CAST\Imaging\CAST-Imaging-Dashboards\logs restapi.log

Each folder is on the machine where that component is installed - see Deployments with several machines. The Gateway, Auth Service and Control Panel write their log directly in the component folder, not in a logs subfolder. The imaging-windows-*.log and service-error.log files in the viewer folder are written by the Windows services that start each viewer component: send them when a viewer service does not start. In releases before 3.6.8-funcrel, the SSO log is CAST-Imaging-SSO-log.txt in the SSO folder.

To send everything from one machine, run the following in PowerShell as an administrator. The services keep their logs open, so the command copies them to a temporary folder first, then compresses the copy into cast-imaging-logs-<machine>.zip on your desktop. Folders that do not exist on the machine are skipped, so you can run the same command on every machine of a distributed deployment:

$dst = "$env:TEMP\cast-imaging-logs"
Remove-Item $dst -Recurse -Force -ErrorAction SilentlyContinue
$src = @{
  "analysis-node" = "$env:ProgramData\CAST\Imaging\CAST-Imaging-Analysis-Node\logs\*"
  "sso"           = "$env:ProgramData\CAST\Imaging\CAST-Imaging-SSO\logs\*"
  "console"       = "$env:ProgramData\CAST\Imaging\CAST-Imaging-Console\logs\*"
  "gateway"       = "$env:ProgramData\CAST\Imaging\CAST-Imaging-Gateway\*.log"
  "auth-service"  = "$env:ProgramData\CAST\Imaging\CAST-Imaging-Auth-Service\*.log"
  "control-panel" = "$env:ProgramData\CAST\Imaging\CAST-Imaging-Control-Panel\*.log"
  "dashboards"    = "$env:ProgramData\CAST\Imaging\CAST-Imaging-Dashboards\logs\*"
  "viewer"        = "$env:ProgramFiles\CAST\Imaging\CAST-Imaging-Viewer\logs\*"
}
foreach ($k in $src.Keys) {
  if (Test-Path $src[$k]) {
    New-Item -ItemType Directory -Path "$dst\$k" -Force | Out-Null
    Copy-Item $src[$k] "$dst\$k" -Recurse -Exclude *.zip
  }
}
Compress-Archive -Path "$dst\*" -DestinationPath "$env:USERPROFILE\Desktop\cast-imaging-logs-$env:COMPUTERNAME.zip" -Force
Remove-Item $dst -Recurse -Force

The command leaves out the <application>.zip files of the viewer folder: these are the exported results sent to the ETL, and can be large. Send them only when Support asks for them.

From disk on Linux via Docker/Podman

Each folder is on the machine where that component is installed - see Deployments with several machines. The logs are written to the host, so you can read them without access to Docker:

Service Default location Files
Node /opt/cast/core/logs analysis-node.log
Control Panel, Gateway, Auth Service, Console, SSO /opt/cast/imaging-services/logs, one subfolder per service: control-panel, gateway, auth-service, console, sso <service>.log, for example gateway.log
Viewer /opt/cast/installation/imaging-viewer/logs imaging-service.log, and access.log for the requests received
ETL /opt/cast/installation/imaging-viewer/logs imaging-etl.log
Imaging API /opt/cast/installation/imaging-viewer/logs imaging-api.log
AI Service /opt/cast/installation/imaging-viewer/logs ai_service.log, agents/imaging-agents.log, graphrag/imaging_graphrag.log
Neo4j /opt/cast/installation/imaging-viewer/logs neo4j.log, debug.log, query.log, security.log
Dashboards /opt/cast/imaging-dashboards/logs restapi.log

The node, imaging-services and dashboards logs roll over daily: the folder holds the current log plus one compressed file per previous day, for example gateway.log.2026-11-10.0.gz. To send everything from one machine, compress the folders that exist on it. On a machine that does not host every component, tar reports the missing folders and compresses the others. Run the command on every machine of a distributed deployment:

$ sudo tar czf cast-imaging-logs-$(hostname).tar.gz /opt/cast/core/logs /opt/cast/imaging-services/logs /opt/cast/installation/imaging-viewer/logs /opt/cast/imaging-dashboards/logs

The imaging-services and dashboards folders can be moved with IMAGING_LOGS_HOST_MACHINE in the component’s .env file - see Docker/Podman file storage locations. The viewer folder is inside the imaging-viewer installation folder, so it moves with the installation.

Container logs (Linux via Docker/Podman)

Every container also writes to its standard output. Support asks for this when a container does not start, or when the problem is not in a log file. Run the following on the machine that hosts the container, as a user that can run Docker commands, replacing <container> with a name from the table:

$ docker logs <container> > <container>.log 2>&1

To save the output of every container on the machine at once:

$ for c in $(docker ps -a --format '{{.Names}}'); do docker logs "$c" > "$c.log" 2>&1; done
Component Containers
imaging-services gateway, console, auth-service, sso-service, control-panel, postgres (embedded PostgreSQL only)
analysis-node analysis-node
imaging-viewer server, etl, imaging_apis, open_ai_manager, neo4j
dashboards dashboards

Use podman instead of docker on Podman. Run docker ps -a to list the containers on the machine, including stopped ones.

Browser logs: console and HAR

When the problem appears in the UI, for example a blank page, an error message or a view that does not load, Support may ask for the browser console or a HAR file. Both are produced by your browser, so no access to the servers is needed. These steps apply to Google Chrome and Microsoft Edge:

  1. Open the page where the problem occurs, and press F12 to open the developer tools.
  2. In the Network tab, select Preserve log, then reproduce the problem.
  3. To save the HAR file, right-click in the list of requests and select Save all as HAR with content.
  4. To save the console, open the Console tab, right-click in it and select Save as.

Installation and update logs

Platform Components Location
Microsoft Windows imaging-services, analysis-node, dashboards %PROGRAMDATA%\CAST\CAST\.install, one file per run named YYYYMMDDHHmm_cast-imaging-[install|update|remove].log
Microsoft Windows imaging-viewer %PROGRAMFILES%\CAST\Imaging\CAST-Imaging-Viewer\logs: imaging-viewer-installer.log and neo4j-installer.log
Linux via Docker/Podman imaging-viewer /opt/cast/installation/imaging-viewer/logs/cast-imaging-install.log
Linux via Docker/Podman Other components No log file: save the console output of the installer or of the update script, and the container logs of any container that does not start

Delivery folder and schema backup

This is not a log, but Support may ask for it to reproduce a problem on their side. It is an application backup that contains, depending on the options you choose:

  • the application schemas *_MNGT, *_LOCAL and *_CENTRAL, in .cssbackup format
  • the application’s delivery folder, which holds the delivered source code, the analysis settings such as exclusion patterns, and the fast scan data

You need the Admin profile or a profile with the Administrator role:

  1. Open the application’s details from the Actions menu, as for the CAST Support section, then expand Backups.
  2. Select BACKUP APPLICATION, then select Backup application schemas, Backup delivery folder, or both, depending on what Support asked for. Select BACKUP.
  3. When the backup is listed, select its download icon. Its Type column shows Schema, Delivery or All.

See Managing application backups for details. Backups are also stored on disk, in the backup folder of the shared common-data folder: <common-data>/backup/<application>/<application>_YYYY-MM-DD.zip.

Other logs

Log Where to find it
PostgreSQL Embedded instance on Linux: /opt/cast/cast-imaging-db-data/data/log, one postgresql-<date>_<time>.log file per day, readable by root only: sudo cp -r /opt/cast/cast-imaging-db-data/data/log ./postgres-logs. Other instances: the log folder of the PostgreSQL data directory
CAST Extend local server docker logs cast_extendproxy - see CAST Extend local server
CAST Imaging MCP Server See the troubleshooting section of the MCP Server installation page for your platform
Kubernetes deployments See Downloading logs