Hi,

The incremental merge is much faster than deep merge because incremental merge does not modify the "base file". For example, if you incremental merge B to A, then pretty much all it does is to append B to the end of A, then append a newer version of the PDF file "root" object at the end of the file. PDF file is designed in such a way that if the same objects are being encountered more than once, then the later one is the effective one. This way the result file will have two "root" objects. One is the one that A already has. Another one is the new one that got appended to the end of the result file. The second one is the one that will be used when a PDF Viewer tries to render the file. The same goes for other resources in the file: for example, if both file uses the same font and the font data are embedded, them in a incremental merge the result file will have two duplicate font data blocks.

Deep merge on the other hand scans the whole document and merge everything as much as possible. In this case there will be a single result root object. The result file will have a completely new layout regenerated and nothing has to match the original file except for the contents should still be the same. A deep merge usually produces smaller and also "cleaner" result file --- the downside is what you have already noticed: it's more CPU and memory intensive than an incremental merge.

Thanks!

Hi,

You can try to set HtmlToPdf.Options.RetrieveNodeText to false and that should reduce memory usage. If that still does not work, you can try to:

1. Reduce the size/complexity of your HTML file;
2. Run your conversion inside a separate AppDomain and then restart that AppDomain whenever you get an out of memory exception;

The reason for #2 is memory fragmentation. After a while even if every chunk of memory that is no longer needed is freed, you will still receive out of memory exception if your usage memory space becomes fragmented and there is no single continuous block that is big enough to satisfy the allocation request. In that case you will receive out of memory exception even if you still have enough available memory --- it's just that they are divided into numerous small pieces. In that case the most effective way is to restart the AppDomain the converter runs.

Thanks!

Hi,

You can try to increase MaxConcurrenTaskCount and see if you get into other problems. If you increase it and get other problems such as OutOfMemoryException, or conversion time out, or simply that it decreases the overall performance rather than increasing it, then you are overloading your system and you should decrease your parallel level (decrease number of worker thread, add wait period between taskes, etc).

Thanks!

Here’s the mail I got recently for my problem
The MEMORY_MANAGEMENT error caused the blue screen and in turn the computer shut itself down to prevent further damage to your computer which in turn caused this error to happen. As the log says "This error could be caused if the system stopped responding, crashed, or lost power unexpectedly" proves that the sole reason of this error was from the MEMORY_MANAGEMENT error and is not the MEMORY_MANAGEMENT error log itself. This error is not as severe as the MEMORY_MANAGEMENT error and you shouldn't need to fix it either.
How To Fix MEMORY_MANAGEMENT – 0x0000001A?