What is APT_CONFIG_FILE ?

 

APT_CONFIG_FILE is an environment variable that points to a configuration file (.apt file) which defines:

• Number of processing nodes

• Node names

• CPU allocation

• Disk resources

• Scratch disk paths

• Resource pools

This file controls parallelism in Data Stage PX (Parallel Extender).A datastage job cannot run without .apt configuration file.

---

Example APT Config File (2-node)

node "node1"
{
  fastname "server1"
  pool "node"
  scratchdisk "/scratch1"
}

node "node2"
{
  fastname "server2"
  pool "node"
  scratchdisk "/scratch2"
}

This means the job will run on 2 nodes, in parallel.

---

Why is APT_CONFIG_FILE Important?

✔ Controls job parallelism

More nodes = higher performance.

✔ Controls resource usage

Disk paths, node pools, etc.

✔ Controls load distribution

E.g., partitioning, sorting, aggregations all depend on node definitions.

---

Where is APT_CONFIG_FILE set?

Usually set in:

• Project → Properties → Environment Variables → APT_CONFIG_FILE

• Or in DS job → Job Properties

• Or inside UNIX shell before execution:

export APT_CONFIG_FILE=/opt/IBM/InformationServer/Server/Config/apt_config.apt

---

Q: What happens if APT_CONFIG_FILE has 4 nodes but your job has only 1 input row?

Job will still run in 4-way parallel, but only 1 node will get the row; others will be idle.

Q: How do you check which config file your job is using?

In Director → Right-click job → View Log → It shows the APT config file name in the startup message.

---