I've been battling this issue for quite sometime now. I'm not sure if this is a ArcGis Online question or an Enterprise question, but I'll make two posts to cover bases. Enterprise and Portal is relativity new to us and we're self taught, learning as we go.
We have an 'Electric Network' database that is edited everyday in the office using ArcMap. We add new accounts, add power lines, poles etc.... We want our workers to have the latest/greatest info in the field, so we have a script that runs each night and exports data from our 'Electric Network' to a SQL Enterprise Geodatabase. I'm not a code guys, but I think the script deletes everything in each layer and appends it back with the new stuff added as well. From that data, I have published Web Layers to our Portal and use these layers to build maps for our workers to use in Field Maps. These layers are for viewing purposes only, no edits will be made to these layers as they will be overwritten each night when the script runs.
Our organization relies heavily on Offline Maps as we don't have good cell service in our area. This is where we run into problems.
Using OFFLINE maps, it is our understanding the Web Layers must have 'Sync' enabled, therefore each layer must have Global IDs. In our first attempt in all this we would dump the data into a 'Versioned Database'. Our issue was the Versioned database became bloated with data in the background tables because each time the script ran, new GLOBAL IDs were created for each record and ESRI saw that as and edit. After about a week of this script running the Versioned database tables (delta/edit/delet tables?) would be bloated (some with close to a million records) and it would gradually slow things down, almost to a complete halt. I couldn't figure out how to clear those records out so I went another direction. I attempted to reconcile/post but the records in the edit/delete/delta tables wouldn't clear out.
We decided to move away from a Versioned database, we didn't think it was necessary as nobody was going to make versions and edit this data, and it was hanging onto all that info/edits from the script running. So we changed the script to dump the data into a Non-Versioned database. All is well, the database no longer gets bloated with 'edits' when the script runs, I view it in SQL and can see the tables aren't being bloated like they were when it was versioned. Great! Now our issue is each time a user Syncs their Offline Map, they data from the Electric Network is duplicated... sync again.. its tripled... sync again its quadrupled. Each time the 'Sync' takes longer and longer, and eventually the offline map starts slowing down, layers wont draw. I've had an ESRI case open on this for weeks and we've yet to figure out what's going on. We check the data source and can't find any duplicated data anywhere... We've deleted the offline map from the phones and redownloaded it and it still comes in duplicated/tripled/quadrupled etc...
I don't really care what type of database we go with, versioned, non-versioned... whatever, I don't care, I just need this to work seamlessly. I can't rush in each morning before the crews go out and make a new Offline map for them to download that day. I'm a self taught Enterprise/Portal guy, there must be something I'm not doing right or have overlooked. I know what's happening, it's because the scripts is making new Global Ids each time it runs, but the Offline Maps don't seem to be letting the old data go
any help on this would be great.
