Having now been running collector within our organisation for the last 4 years I have come across several issues with managing the ever increasing replicas generated from substantial use. One of our databases has over 400 replicas due to the volume of users and number of maps downloaded.
There are several scenarios in which orphaned replicas can be created in the offline workflows:
Recently I had a scenario in which a database had blown out substantially in size, after detailed investigation there were several replicas which had been orphaned for some time and were storing a large amount of changes. Currently the tools are problematic to identify which replicas we can safely remove and which need to be kept.
doing this manually is very time consuming and has dangers due to the interface in which you can remove a replica is not able to have a filter applied, and crashes when you sort the table more than once.
You can access using python the last sent and last received dates for a replica but you can't access when a replica was created? Neither can you un-register a replica using python, although I suspect this is a good thing as this has large disaster potential. We do need a better way to mange this potential mine field of performance and storage liability better.
Personally I would like a means to use the properties you can get from arcpy.da.ListReplicas to include the created date and be able to use the Replica Class properties in other tools or as a means to filter the replicas in a more stable dialog window. Potentially an ability to "TAG" a replica with the results of a python script would be incredibly helpful.
This BUG with the Replica Manager button has been consistent since 10.3 when I first found it. Still waiting for a fix other than don;t click on the column title more than once.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.