Troubleshoot IBM i journals and receivers

Read the journal position of the pipeline from the Core Hub REST API. <pipelineId> is the identifier shown in the pipeline URL (for example 9fc6ac3b) and <token> is a valid API token (see Core Hub APIs):

curl -k -H "Authorization: Bearer <token>" \
  https://<core-hub-host>/pipelines/<pipelineId>/read-checkpoints

In the Control Plane, open Query Studio;

In SQL-capable Agent, select your pipeline and then its (SOURCE) agent;

Run the query that lists the receivers of your journal library, oldest first:

SELECT JOURNAL_RECEIVER_NAME AS RECEIVER, STATUS,
       FIRST_SEQUENCE_NUMBER AS FIRST_SEQ,
       LAST_SEQUENCE_NUMBER AS LAST_SEQ,
       NEXT_JOURNAL_RECEIVER AS NEXT_RECEIVER
FROM QSYS2.JOURNAL_RECEIVER_INFO
WHERE JOURNAL_RECEIVER_LIBRARY = 'GSDOCSJRN'
ORDER BY CAST(FIRST_SEQUENCE_NUMBER AS BIGINT)

Start the entity (Start sync, then Start entity). If the table is not journaled for CDC, the start fails and a red notification appears;

Read the notification. Its title is Missing Journal Table Definition and its text starts with Missing journal configuration for entity <name> and table <LIBRARY.TABLE>. Sync cannot be started. Only snapshot can be executed. The entity stays in the Hold group;

Ask the IBM i administrator for the journaling of the file:

SELECT OBJECT_LIBRARY, OBJECT_NAME, JOURNAL_LIBRARY, JOURNAL_NAME, IMAGES
FROM QSYS2.JOURNALED_OBJECTS
WHERE OBJECT_LIBRARY = 'GSDOCSJRN' AND OBJECT_NAME = 'NOJRN'

If the query returns no row, the file is not journaled. Start journaling with both images:

STRJRNPF FILE(GSDOCSJRN/NOJRN) JRN(GSDOCSJRN/JRNDEMO) IMAGES(*BOTH) OMTJRNE(*OPNCLO)

If IMAGES is *AFTER, end journaling and start it again with both images:

ENDJRNPF FILE(GSDOCSJRN/AFTERONLY) JRN(GSDOCSJRN/JRNDEMO)
STRJRNPF FILE(GSDOCSJRN/AFTERONLY) JRN(GSDOCSJRN/JRNDEMO) IMAGES(*BOTH) OMTJRNE(*OPNCLO)

Start the entity again with Full copy, so that rows changed while the file was not journaled reach the target (see [resync-entities]).

With the receiver deleted, start the entity with Where it stopped (the default for an entity that has run);

Wait for the first journal read, then open the notification Journal checkpoint outdated.

Pause the entity (Stop sync, then Pause entity);

Select Start sync and, in Start from, select Full copy;

In Data already in the target, select Keep and merge if no row was deleted on the IBM i in the meantime, or Empty the table first if rows may have been deleted;

When you empty the table, tick I understand that rows in the target will be deleted. Start entity stays disabled until you tick it. If you do not see the box, scroll the dialog;

Select Start entity.

Select Settings (the gear icon) and then Danger zone;

In Pipeline checkpoint reset, select the pipeline in Select a pipeline;

Select Reset pipeline checkpoint;

Read the dialog Confirm Reset and select Confirm.

Look at how the IBM i names the next receiver (CHGJRN …​ JRNRCV(*GEN) generates it). With a short name such as JRCV9999 the number simply gets longer (JRCV10000). With a name that already has ten characters, such as JRCVAA9999, the number wraps to JRCVAA0000;

Run the receiver query of step 1 and check that every receiver except the attached one names its successor, also across the wrap.

Pause the entity;

In the entity row, select Options (the three-dot icon) and then Edit settings;

On the Source tab, switch Use Dedicated Journal Reader to On;

Close the dialog and select Confirm and continue;

Start the entity with Where it stopped.