This thread is for posting various lessons learned that customers have surrounding the use of Landmark Replication Sets.
From a customer on the call:
One gotcha that we have experienced in the past is that, while the limit for FTP file size of replication sets is 10MB, ION data lake chunks the files even further down to 5MB. This can be an issue if you taking advantage of the python scripting ability within ION to manipulate the data from the replication, especially if you are then outputting said data to an SFTP location.
From another customer:
I found that replication sets should have unique names across environments. Had a replication set HCM.TEST and an FSM.TEST. Both were scheduled for midnight. HCM.TEST was taking 15 minutes, FSM.TEST was taking 8 hours. HCM.TEST was being locked until FSM.TEST finished.
You can turn on a feature which appends a unique replication set id to each replication. That way the output files don't overwrite.
If anyone is interested in that feature, it's a configuration parameter. The configuration parameters you can use are export.file.name.versioning or export.file.name.versioning.RepSetName. For more information, this is documented at docs.infor.com > Technology > Landmark. Search for 'Configuring file names for replication output'.
