Memory issue with ‘Extract from XML’ node

Questions
,
1

Hello everyone.
I have created a workflow to dynamically import XML files into Postgress. Everything works great. N8N is great, however I am having problems with large files.

I downloaded a zip file and unzipped it. The extracted XML file is 133 MB and when I use the ‘Extract from XML’ node I have a memory problem: “There might not be enough memory to finish the execution.”

Do you have any solutions to solve this problem?
Is it possible to break the XML file into smaller blocks? Any other ideas?

Once I have extracted the XML file I convert it to JSON, so if there was also a way to transform the XML file contained in the ZIp file directly into Json I could skip a step.

Thanks in advance to anyone who can help me
Davide

2

It looks like your topic is missing some important information. Could you provide the following if applicable.

  • n8n version:
  • Database (default: SQLite):
  • n8n EXECUTIONS_PROCESS setting (default: own, main):
  • Running n8n via (Docker, npm, n8n cloud, desktop app):
  • Operating system:
3

Hey @Davide_Wonder,

Welcome to the community :tada:

How are you running n8n and how complex is the XML file?

4

Hi Jon,
thank you for your interest.

I’m using the trial cloud version and I’am thinking of buying the “starter” or “pro” cloud plan to get started.

I have seen that other versions have more RAM but I would like to solve the memory problems in another way.

No, the XML file is not complex, it only has so many records and therefore weighs several MB.

XML example:

<?xml version="1.0"?>
<ALLRECORDSDATASET>
  <RECORD>
    <FDI_SB00>0119032944011</FDI_SB00>
    <FDI_9ARG>01</FDI_9ARG>
    <FDI_9AAS>19</FDI_9AAS>
    <FDI_0001>032944011</FDI_0001>
    <FDI_9222>0.000</FDI_9222>
    <FDI_9190>N</FDI_9190>
    <FDI_9079>N</FDI_9079>
    <FDI_9181>N</FDI_9181>
    <FDI_9183>N</FDI_9183>
    <FDI_9187>N</FDI_9187>
    <FDI_9185>N</FDI_9185>
    <FDI_9189>N</FDI_9189>
  </RECORD>
  <RECORD>
....
  </RECORD>
......
</ALLRECORDSDATASET>

Thanks for your help
Davide