Ok, I have read post where tables can have a lot of data and cause the serializing of windows to take an extended period of time. I have gotten rid of all of my tables but it still is taking over 5 minutes to complete serializing. Am I missing a step to get this moving quickly again?
It’s not so much the tables, but dataset properties that hold a large BasicDataset. Current versions of Ignition have a checkbox in query bindings that let you throw away the data on serialization. This works by returning an alternate implementation of the Dataset interface. It doesn’t help you when you create datasets in scripting.
I may have been the first to really suffer from this problem: I created such transient dataset types some years ago for my Time Series Caching Module, as it produces enormous datasets and the serialization was killing my projects. I recently moved the core class for that to my Simulation Aids Module for wider (free) availability, including in older versions of Ignition (back to v7.7.8).
Yeah, I just cannot fins where those bindings are. I have gone through all the tags being used and they are not using any bindings.
Bindings within tags (expression tags) are not relevant. It is components in windows that have large datasets on properties.
If you right click the window object while holding down the shift key, there will be an option to ‘Copy XML to Clipboard’ - this will be the Java serialization of the state of the Window and everything on it, and you should be able to find your large datasets that way.
2 Likes
Perfect. I copied all of this to clipboard and I do see huge amounts of data in a couple sections. How can I tell where they are located? I am reading the info before the data occurs and it is really not saying much.