Reply to thread

Hundreds of nodes really shouldn't be a huge problem. If you had 300 nodes in total and even if each of those led to a total of 10 entities (2 is closer to realistic for the node and type-specific record), you'd have instantiated 3000 entities. If we assume each takes up 10KB of memory (in a very quick and non-scientific test, I had them taking about 9KB each), you're in the vicinity of 30MB of memory and well short of the standard 128MB limit.

Thread deletion is handled by a job so that shouldn't trigger much additional memory (though the job can be kicked off immediately).

So it does seem like something else is likely taking up a fair amount of memory. It's worth noting that we do try to limit the size of the query log once it gets over 150 queries, but it's still something that only exists in debug mode and there's no guarantee of how much memory it might take up there.

If we start hearing about this on production forums that might be something to look into further.