Skip to content

Re-indexing ​

This process brings the indexing in the archive up to the state of SAP. When a business partner, an item or a document changes, the archive still holds the old index values – this run catches them up.

It concerns two kinds of record: master data (business partners, items) and documents. Both are treated differently, and that difference explains most of the questions about it.

When it runs ​

A timer of type A (also written INDEX) or OVERWRITE, set up under timers. The difference between the two is right below.

Like every ECM process it needs the user running it to hold the ECM-Connect permission.

What happens ​

  1. The configured indexings are walked through, per SAP object.
  2. For each record it is decided whether it has to be handled at all.
  3. For the records to be handled the index values are determined again and written to the archive.
  4. Every successfully indexed master record is noted so the next run can skip it.

When a record is skipped ​

This is the heart of the process – and the reason a run often "does nothing":

CaseWhat happens
Master record already indexed, nothing suggests a changeskipped
Master record already indexed but inside the change windowindexed again – a change counts for more than "done before"
Documenthas its own check through the link table, independent of master data
Run of type OVERWRITEskips nothing – everything is written anew

When everything should be written anew

An ordinary run works sparingly: whatever is indexed already and has not changed is left alone. When the whole stock should be indexed again – for instance after you changed a field mapping – you need a timer of type OVERWRITE. It takes longer but leaves nothing out.

How it is recognised what is already indexed

For master data a single query determines which records are already in the archive. Documents use the link table for it. Both routes lead to the same goal but are separate – which is why documents may be caught up while master data is not, or the other way round.

What steers it ​

WhatWhere
Timer, type A or OVERWRITEtimers
Which objects are indexedindex and master data
Which fields go into the archivefield mappings
SAP objectsSAP objects
Watching a runECM monitor

What can go wrong ​

The run reports success but the archive still holds the old value ​

The record was skipped because it counts as indexed already and nothing suggests a change. Two routes:

  • If the SAP record did change, the change window should apply – check whether one is configured for this indexing.
  • If you changed a field mapping, only a run of type OVERWRITE helps: the change is in the configuration, not in the record, and cannot be seen on it.

Master data is caught up, documents are not (or the other way round) ​

Both have separate checks. That one side works while the other does not is no contradiction – see the note above.

An OVERWRITE run takes a very long time ​

That is to be expected: it skips nothing. Schedule it outside working hours and do not let it run permanently – type A is the one meant for normal operation.

Documentation – NC-Solutions