Fatal error - memory issue but there is enough memory

General
1

Hi, can anyone help me discover the reason of fatal-error on this part of the log:

And is there a way to automatically start node-red after a fatal-error? we use this as a school project for monitoring several spaces and do a research on the data, but this becomes a problem if there is for hours data missing.

Thanks!!

We use a simple MQTT node that receives this every 2 seconds:
image

And we use the buienradar node. interval 5 minutes.

It runs on a VPS with debian, 50GB SSD and 1GB memory

Statistics show that there is a large increase of HDD and CPU usages starting on 4:42, the fatal error is on 5:41

FATAL ERROR: Committing semi space failed. Allocation failed - process out of memory
1: node::Abort() [node-red]
2: 0x8cd14c [node-red]
3: v8::Utils::ReportOOMFailure(char const*, bool) [node-red]
4: v8::internal::V8::FatalProcessOutOfMemory(char const*, bool) [node-red]
5:
v8::internal::Heap::PerformGarbageCollection(v8::internal::GarbageCollector,
v8::GCCallbackFlags) [node-red]
6: v8::internal::Heap::CollectGarbage(v8::internal::AllocationSpace,
v8::internal::GarbageCollectionReason, v8::GCCallbackFlags) [node-red]
7: v8::internal::factory::NewFillerObject(int, bool,
v8::internal::AllocationSpace) [node-red]
8: v8::internal::Runtime_AllocateInNewSpace(int,
v8::internal::Object**, v8::internal::Isolate*) [node-red]
9: 0x2349e67042fd
Afgebroken

2

So your are still having memory problems like your "solved" post...

You may have temporarily masked the problem by changing poll rate to 5 seconds but an underlying issue exists.

As I said in the previous post, it looks like xpath module inside that node is being overwhelmed. Raise an issue with the author of that node.

3

Thanks, i did'nt see the buienradar node on the error_log like on the other post. so i thought that is not causing the problem but something else is causing it. I already did raised an issue.

4

Maybe worth exploring an alternative weather node.

https://flows.nodered.org/?term=weather&type=node&num_pages=1

5

Have you checked how often their data is really updated? I can't read the language I'm afraid. Most weather API's update hourly or even less. So you may be creating an issue for yourself.

Also, without seeing the flow, we can't tell what else is going on.

Getting an OOM error on that spec certainly seems to indicate some kind of serious memory allocation happening faster than the GC can cope with.

6

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.