Mismatched Activation Group

117 Views

Prev
Next

Here's a realistic scenario where a wrong or mismatched activation group causes a runtime error like:

RNX1217 — Program or service program not found
RNX1216 — Level check or file already open error

…and how changing the activation group resolves it.

🧨 Scenario: Activation Group Mismatch Causes Crash

You have:

  • A main RPG program (CUSTOMERPGM) compiled with ACTGRP('CUSTGRP')

  • A service program (CUSTSRVPGM) compiled with ACTGRP(*CALLER)

  • A batch CL job or API call that runs CUSTOMERPGM twice

🔁 Problem Flow

First Call:

CALL PGM(CUSTOMERPGM)

  • Activation group CUSTGRP is created

  • CUSTOMERPGM runs

  • CUSTSRVPGM runs inside CUSTGRP (because it uses *CALLER)

  • All files and globals stay open

Second Call (same job):

CALL PGM(CUSTOMERPGM)

Now the system tries to:

  • Load CUSTOMERPGM into CUSTGRP again

  • Rebind CUSTSRVPGM — but it's already loaded

  • Reopen files already open → file already open error or level check error

💣 Result: Crash

  • You get:

    • RNX1216 (file open or level check)

    • Or RNX1217 (binding issue)

  • Especially if files changed between runs or you recompiled only one component

✅ Fix: Change ACTGRP(*NEW) in Main Program

In CUSTOMERPGM:

ctl-opt actgrp(*new);

Now, every call to the program:

  • Creates a new activation group

  • Loads fresh copies of service programs and files

  • No residual memory or bindings from the previous call

✅ Alternative Fix: Use RCLACTGRP

If you want to keep using a named group (CUSTGRP) but reset it before rerun:

RCLACTGRP ACTGRP(CUSTGRP)
CALL PGM(CUSTOMERPGM)

🧠 Takeaway

Problem Fix
Reusing a named group causes errors Use *NEW to isolate memory
Files or service programs stay loaded Use *CALLER and control cleanup properly
Need to fully reset environment Use RCLACTGRP to clear memory

Would you like a CL wrapper that auto-recalls RCLACTGRP before each call to ensure fresh execution — especially useful in API or batch chains?

Prev
Next
Save ()

Post Comments