It can be reproduced in production with customer’s real data. But we are still trying to reproduce the memory issue on local. I am not able to provide a testing code for you.
From the output of analyzing the OOM memory dump file, we found the number of objects of two Aspose classes is far bigger than the data we were rendering. (The data size is very small, 13 rows in total).
And we can see from the following screenshots, the 99.83% memory usage is by two classes(both are Aspose classes), they are com.aspose.words.Run and com.aspose.words.FieldSeparator.
Do you happen to know which improper usage would cause this problem? Btw, we are using Aspose 15 in Java.
image.png (322.0 KB)