You need at least one active execution server to run tasks.

In this page

• License models
• Quick start
• Server pages
• Create/Edit: field reference
• Operational checks
• Multi-servers notes (load balancing)
• Troubleshooting

License models

• MONO — single node (control + execution).
  • See: Mono-server installation
• MULTI — one control node (Main server) and one or more Execution servers (slaves) for load balancing and redundancy.
  • See: Multi-servers installation

In both cases, tasks only run on nodes flagged Is Execution Server.

Quick start

MONO license (single node)

1. Install EAC on the host.
2. Set the server identity in the host config: EAC_SERVER_CODE (e.g., master).
3. In Servers → New, register the node and tick Is Execution Server and Enabled.
4. Schedule:
   • Heartbeat (on this node by default):
     * * * * * php bin/console eac:server-status –every=3
     30 1 * * * php bin/console eac:server-status:purge --days=4
   • Scheduler loop (on this node by default):
     * * * * * php bin/console eac:scheduler (and set SCHEDULE_ENABLED=true)

MULTI license (control + slaves)

1. Install EAC on each node.
2. On each node, set a unique EAC_SERVER_CODE (e.g., master, slave-1, slave-2).
3. Register every node in Servers. On the control node, tick Is Main Server. On nodes that will run jobs, tick Is Execution Server. Tick Enabled wherever applicable.
4. Schedule on every node:
   • Heartbeat:
     * * * * * php bin/console eac:server-status --every=3
30 1 * * * php bin/console eac:server-status:purge --days=4
5. Schedule the scheduler loop on the Main server (recommended):
   • * * * * * php bin/console eac:scheduler with SCHEDULE_ENABLED=true
   • (Optional HA) keep a standby scheduler on another node with SCHEDULE_ENABLED=false, and flip it only if needed.
6. Ensure job artifacts are available on all execution servers (see Replication checks in Useful commands or:
   * * * * * php bin/console eac:check-jobversion-deployment
0 2 * * * php bin/console eac:check-jobversion-deployment --force)

Server pages

• List — high-level view: display name, code, role badges (Current, Master), enabled state, and quick actions.
• Show — server details + Last executions on this server. CPU/Memory graphs appear once heartbeats are flowing. If you see “No status available”,
  • your heartbeat isn’t configured or isn’t reaching the app.
  • your last executions exceed time defined in fine last server status
• Create/Edit — the registration form (also used for updates). This is where you define Main/Execution roles.

Create/Edit: field reference

On the Show page you’ll also see: Active (heartbeat seen recently), EAC version, OS, and CPU/Memory charts.

Operational checks

• Heartbeat present? Status appears and updates every minute once eac:server-status is scheduled.
• Execution capable? At least one server must be Enabled and Is Execution Server.
• Scheduler running? Ensure the scheduler loop is scheduled and SCHEDULE_ENABLED=true on the node where it runs.
• Artifacts available? In MULTI, verify replication on all execution servers (see replication checks above).

Multi-server notes (load balancing)

• With MULTI, tasks will run on any available Execution Server.
• Keep replication healthy so the required job versions exist on all candidate nodes before triggers fire.
• Use the Server Show page to see which server actually executed each run and to spot imbalances or capacity issues (CPU/Memory charts).

Troubleshooting

• “No status available” on Show page
  • Heartbeat cron missing or failing; wrong Code (doesn’t match EAC_SERVER_CODE); firewall/TLS blocking Host:Port.
• No tasks start
  • No server with Is Execution Server + Enabled; scheduler loop not running or SCHEDULE_ENABLED=false; task blocked by Ack; schedule not in the look-ahead window (“Active now” false).
• Executions always on one node
  • Other nodes not marked Is Execution Server, or their artifacts aren’t replicated.
• Version mismatch
  • Different EAC version or missing job version on some nodes; run replication checks.

See also

• Useful commands & operations
• Scheduler · Good use of context · Purge · Ack

L’article Execution servers est apparu en premier sur ETL administration console.

• How to use settings
• Scheduling & System
• Mail settings
• Database
• ETL Table Settings
• Examples

Use the Settings page to configure scheduling, system paths, email, databases, and ETL logging.

• Scope & precedence
  • .env = baseline (read-only in the UI).
  • .env.local = per-server overrides written by the UI. Anything set here overrides .env on that server only.
• Single-server: edits apply to the only node.
• Multi-server: you edit the current server. all changes are applied automatically to your servers.

After saving, EAC writes the target server’s ./shared/.env.local (or equivalent) and refreshes configuration. Some changes (paths/scheduling) may require worker/cron reload.

How to edit settings (UI)

1. Sign in as Admin.
2. Go to Settings.
3. For each section, change Updatable (.env.local) values. Present value (.env) is informational.
4. Save. Confirm the success banner. If prompted, restart workers/cron.

To remove an override, clear the field and save (the app falls back to .env).

Field reference (mapping → env keys)

Below, the UI label is mapped to the environment variable actually written in .env.local. Defaults are indicative—your build may ship different ones.

Scheduling & System

Permissions tip: ensure the service user (e.g., eac:www-data) owns these paths and they exist.

Mail settings

You can either fill the discrete fields or paste a full DSN (advanced). The UI composes MAILER_DSN from individual fields.

Resulting DSN written:

MAILER_DSN="smtp://<user>:<pass>@<host>:<port>?<options>"
# Example
MAILER_DSN="smtp://mylogin%40acme.com:My%3APass@smtp.acme.com:587?encryption=tls&auth_mode=login"

Percent-encoding for credentials in URLs

Only encode the username and password parts of the URL (not the whole URL).

Format reminder:

scheme://USERNAME:PASSWORD@HOST:PORT/DB?options

Reserved characters (what they mean & why to encode)

• @ → %40
  Meaning in URL: separates USERNAME:PASSWORD from HOST.
  Encode when: @ appears inside your username or password, otherwise it will be read as the end of credentials.
• : → %3A
  Meaning in URL: separates USERNAME and PASSWORD.
  Encode when: : is part of the username or password value, or it will split the field.
• / → %2F
  Meaning in URL: path separator after the host.
  Encode when: / appears inside credentials, or it will be mistaken for a path delimiter.
• & → %26
  Meaning in URL: separates query parameters (after ?).
  Encode when: & appears in credentials, or it’ll be parsed as a new option.
• % → %25
  Meaning in URL: introduces a percent-encoded byte.
  Encode when: % appears literally in credentials—encode it first to avoid corrupting other encodings.

Other common ones you might hit: + → %2B, # → %23, ? → %3F.

Do & Don’t

• Encode only the username/password segments.
• Keep separators (://, @, /, ?, &) unencoded outside credentials.
• Don’t double-encode (e.g., don’t turn %40 into %2540).

Examples

Database URL (PostgreSQL)

Password contains pa:ss@/word&%

postgresql://eac_usr:pa%3Ass%40%2Fword%26%25@db:5432/eac_db?serverVersion=16&charset=utf8

Mailer DSN (SMTP)

Username alerts@company.com, password P@ss:W/rd&2025

smtp://alerts%40company.com:P%40ss%3AW%2Frd%262025@mail.example.com:587?encryption=tls

Quick checklist: If your credential has @ : / & % + # ? → encode those characters before placing it in the URL.

Database (application)

The form composes DATABASE_URL.

Resulting URL:

DATABASE_URL="postgresql://eac_usr:pass@db.acme.local:5432/eac_db?serverVersion=16&charset=utf8"

ETL Table Settings (optional: stats & metrics)

Resulting URL:

DATABASE_URL_ETL_LOG="postgresql://eac_usr:pass@eac_master-postres:5432/eac_db?charset=utf8&serverVersion=16.0.0"

Examples

Single-server, basic mail + Postgres

APP_ENABLE_SCHEDULING=true
APP_BASE_URI="https://eac.local:4443/"

MAILER_DSN="smtp://eac%40cidwe.com:My%3APassword@mail.acme.com:587?encryption=tls&auth_mode=login"
APP_MAILER_SENDER="eac@cidwe.com"
APP_MAILER_DEFAULT_RECIPIENTS="teameac@cidwe.com"
APP_ENABLE_TASK_MAILER=true

DATABASE_URL="postgresql://eac_usr:pass@db:5432/eac_db?serverVersion=16&charset=utf8"

Multi-server, ETL logs on separate DB

APP_ENABLE_ETL_LOGS=true
ETL_STAT_TABLE_NAME=talend_stat
ETL_MET_TABLE_NAME=talend_met
ETL_LOG_TABLE_NAME=talend_log
DATABASE_URL_ETL_LOG="postgresql://etl_ro:pass@etl-db:5432/etl_logs?serverVersion=16&charset=utf8"

L’article Settings est apparu en premier sur ETL administration console.

Note: “Server” refers to one running EAC application instance. If you deploy multiple nodes (e.g., behind a load balancer), you need a Multi-Server license.

Entering a license in the UI

1. Sign in with an Admin account.
2. Go to Admin → Settings → License (menu label may be System → License depending on version).
3. Choose one of:
   • Paste license key: paste the full text.
   • Enter the company name attached to the license
   • Enter the license type code
   • finally select the license server mode (MONO, MULTI)
4. Click Validate & Save.
5. Confirm the beginning date and the expiration date.

If validation fails, contact us.

Renewals and switching plans

• Monthly → Annual (or reverse): enter the new license key the same way and Validate & Save. The UI will update the edition/cycle immediately.
• Single-Server → Multi-Server: apply the new key, then restart additional nodes as needed.
• Expired license: apply the renewed key; the status will switch back to Valid once saved.

Troubleshooting

Tips

• If you use copy/paste, ensure there are no hidden characters (paste into a plain-text editor first).
• For file upload, verify the file name and extension are unchanged.

L’article Licenses est apparu en premier sur ETL administration console.

When you install your web server (Apache or Nginx), the default service user is usually www-data.
For better security and maintainability, it is recommended to create a dedicated user named eac, with www-data as its default group.
This user will be responsible for deploying, executing, and managing the EAC application.

Default application login (EAC UI)

When EAC is installed, a default application user exists:

login: eac@cidwe.com
password: eac

Important: This account is for initial access only. Change the password immediately or disable/remove the user after creating your own admin account.

Suggested actions

• Change password from the UI right after first login
• Create a new admin user, verify access, then disable or delete eac@cidwe.com.

It is strongly recommended to remove or disable this user or at least to change its password for security reason.

1. Create the eac user

sudo useradd -m -g www-data -s /bin/bash eac

2. Set a password for the eac user

sudo passwd eac

3. Verify the user creation

id eac

Creating users from the UI

1. Go to Users.
2. Click Add.
3. Fill the form:
   • Email (required)
   • First name, Last name (optional but recommended)
   • Password and Repeat Password
   • Roles (multi-select ctrl+click to select more than one):
     • Admin – full access to the UI and settings
     • API – programmatic/API access (use for integrations)
   • Is active – must be checked for the account to be usable
4. Click Save.

Configure Default Permissions

To ensure proper file and folder ownership (replace /var/www/html with your actual installation path):

sudo chown -R eac:www-data /var/www/html
sudo find /var/www/html -type d -exec chmod g+rwx,g+s {} \; -o -type f -exec chmod g+rw {} \;
sudo find /var/www/html -type d -exec chmod 2775 {} \;
sudo find /var/www/html -type f -name "*.sh" -exec chmod 774 {} \;

Notes

You can re-run the above four commands at any time if you accidentally change permissions or encounter issues.

L’article Users rights management est apparu en premier sur ETL administration console.

1. System monitoring — health of your EAC nodes (CPU %, memory %, last heartbeat) and quick KPIs (error executions per day/week/month).
   • 
2. Job monitoring — per-execution logs (Execution logs tab) and optional ETL logs (stats/metrics) collected from your ETL logging database.
   • 

more info on cpu per ping click on the graph a modal will appear with detail

Permissions: You need the Admin role to configure monitoring.

System monitoring (nodes & KPIs)

What you see

• Dashboard: one card per node (master/slaves) with CPU % and memory % sparklines, plus Error Executions gauges (day/week/month).
• Server detail: node status (Active ✓ / Not an execution server), EAC version, and time-series of CPU/memory.

How heartbeats work

Each node periodically sends a status sample (heartbeat) using the console command:

php bin/console eac:server-status --every=3

Schedule it on every EAC node.

Cron example (bare metal / VM)

Run every minute (samples every 3s for ~1 minute):

* * * * * cd /var/www/html/eac && /usr/bin/php bin/console eac:server-status --every=3 --env=prod >/dev/null 2>&1

Troubleshooting (system monitoring)

• “Last status received … hours ago”
  Cron/timer not running or wrong working directory/PHP path. Run once manually: php bin/console eac:server-status --every=3 --env=prod -vv
• Graphs flat at 0%
  The command runs but sampling interval too short. Use --every=3 (or higher).
• Time drift / weird “ago” labels
  Synchronize time (NTP) and verify server timezones.
• Permission denied writing logs
  Ensure the app dir is owned by your service user/group (see “Server hardening” section of the docs).

Job monitoring — Execution logs

Every task shows a Execution logs tab with the full runtime output (enqueue → shell/script path → PID → console lines → success/error).
Use it to:

• confirm the script invoked (*_run.sh path)
• check the PID and parameter values
• see final status (“Task completed successfully” or errors)

Copy/paste the relevant block when opening support tickets.

Job monitoring — ETL logs (stats & metrics)

EAC can display ETL-level stats and metrics (duration, rows, components, etc.) if your ETL tool writes logs to a database (Talend standard tables supported).

Enable ETL logs

1. Enable logging in your ETL studio
   Example (Talend): File → Project Settings → Job parameters → “Stats and logs” and select database as target.
2. Enable ETL logs in EAC
   In ./shared/.env.local: APP_ENABLE_ETL_LOGS=true
3. Point EAC to the ETL log database # MySQL example DATABASE_URL_ETL_LOG="mysql://user:pass@db-host:3306/etl_logs?charset=utf8mb4" # PostgreSQL example # DATABASE_URL_ETL_LOG="postgresql://user:pass@db-host:5432/etl_logs"
4. Declare the table names (adapt to your ETL logging schema): ETL_STAT_TABLE_NAME=etl_stat ETL_MET_TABLE_NAME=etl_meter ETL_LOG_TABLE_NAME=etl_log
5. Link ETL execution to the EAC task
   In your main ETL job, print the root PID marker once (Java example; use any component that writes to stdout): System.out.println("@@eac@@root_pid=" + rootPid); EAC uses this marker to correlate the ETL run with the task execution.

Viewing ETL logs

Open a task → ETL logs tab. You’ll see the structured stats/metrics pulled from your ETL log DB alongside the task run.

Troubleshooting (ETL logs)

• No data in ETL tab
  • APP_ENABLE_ETL_LOGS not true, or missing DATABASE_URL_ETL_LOG.
  • Wrong table names; verify ETL_*_TABLE_NAME values.
  • ETL studio still writing logs to files, not DB.
• Can’t connect to ETL DB
  • Check credentials/host/firewall.
  • Test with a quick CLI: php -r 'new PDO(getenv("DATABASE_URL_ETL_LOG")); echo "OK\n";'
• Runs aren’t correlated
  • Ensure the @@eac@@root_pid= line prints once per run and reaches stdout (not swallowed by log frameworks).

Operational tips

Rotate and archive ETL log tables regularly to keep dashboards snappy.application logs and ETL logs

Keep one heartbeat per node (even non-execution nodes) for complete visibility.

In multi-server setups, verify the master and all slaves show recent heartbeats before troubleshooting ETL.

Use separate credentials for the ETL logs database (read-only if possible).

L’article Monitoring est apparu en premier sur ETL administration console.

Check PHP CLI timezone

php -r "echo date_default_timezone_get(), PHP_EOL;"

Per-server identity (required in multi-server)
In each server’s .env.local:

APP_SERVER_NAME=web-1 # e.g. web-1, worker-a, fr-par-02

Multi-server health & replication

Crontab (on every server)

# Report this server's status every minute (internally every 3 min)
* * * * * php /path/to/app/bin/console eac:server-status --every=3

# Daily purge of old server statuses (keep last 4 days)
30 1 * * * php /path/to/app/bin/console eac:server-status:purge --days=4

Job version replication checks

# Incremental check (new/different only) — runs every minute
* * * * * php /path/to/app/bin/console eac:check-jobversion-deployment

# Full comparison once per day (force a full pass)
0 2 * * * php /path/to/app/bin/console eac:check-jobversion-deployment --force

What they do

• eac:server-status: pushes a heartbeat for each node so the UI can display live server health.
• eac:server-status:purge: cleans old heartbeats.
• eac:check-jobversion-deployment: verifies that ETL packages are replicated on all slave servers (incremental or full with --force).

Scheduler (EAC built-in)

Resolution: supports sub-minute plans (down to seconds), unlike system crontab.
Activation: a plan can be switched on/off from the UI or bounded by start/end dates.

Run the scheduler loop

# Launch the scheduler engine every minute
* * * * * php /path/to/app/bin/console eac:scheduler

Gate via environment variable

• Tasks are triggered only if: SCHEDULE_ENABLED=true
• The Start/Stop Scheduler button in the UI flips this variable instantly (no redeploy, no crontab change).

Tip: In the task list, the Active now indicator turns true only when a run is due within the look-ahead window. A future schedule (e.g., Friday) will show false until the due time approaches.

Background process hygiene

• Check & clean background tasks # Example: sanity check every 5 minutes */5 * * * * php /path/to/app/bin/console eac:check-background-task Kills stale background jobs when necessary.
• Upgrade helper (1.6 → 1.7) php bin/console eac:clean-jobversion

ETL logs, stats & metrics

Configure your ETL studio

File > Change project properties > Job parameters > Stat and logs

Choose “save logs in database”.

Enable logs in EAC

In .env.local:

APP_ENABLE_ETL_LOGS=true

# Connection string to the ETL logs database (MySQL, Postgres, Oracle, SQLite, MariaDB supported)
# Examples:
# DATABASE_URL_ETL_LOG="mysql://user:pass@127.0.0.1:3306/etl_logs?charset=utf8mb4"
# DATABASE_URL_ETL_LOG="postgresql://user:pass@127.0.0.1:5432/etl_logs"
DATABASE_URL_ETL_LOG="postgresql://user:pass@host:5432/etl_logs"

# Table names
ETL_STAT_TABLE_NAME=etl_stats
ETL_MET_TABLE_NAME=etl_metrics
ETL_LOG_TABLE_NAME=etl_logs

Emit the root PID from your ETL job

In your main job (e.g., a tJava):

System.out.println("@@eac@@root_pid=" + rootPid);

ETL log purging (keep storage & UI fast)

You can trigger purges manually in the task page or schedule them (Task details → Cron scheduling → Add scheduler, Task type = purge).

Modes

• By date: delete logs before a fixed date (e.g., “< 2025-07-01”).
• By history: keep only the last X runs (e.g., last 100).
• By duration: delete runs faster than Y seconds with status Success/Warning (e.g., < 30s).
• Multiple purges can coexist (e.g., keep last 100 and also drop very short runs).

One-glance crontab template

# --- EAC core ---
* * * * * php /path/to/app/bin/console eac:scheduler
* * * * * php /path/to/app/bin/console eac:server-status --every=3
30 1 * * * php /path/to/app/bin/console eac:server-status:purge --days=4

# --- Replication checks ---
* * * * * php /path/to/app/bin/console eac:check-jobversion-deployment
0 2 * * * php /path/to/app/bin/console eac:check-jobversion-deployment --force

# --- Background hygiene ---
*/5 * * * * php /path/to/app/bin/console eac:check-background-task

Replace /path/to/app with your actual project root (the path that contains bin/console).

L’article Useful commands est apparu en premier sur ETL administration console.