Memory crashes updating Viewports

[Assemblage]

[so as not to parachute onto Kens's "Minimum RAM memory" thread I've opened a fresh one here]

When updating hidden line elevations, axo's, and section VPs in sheet layers I get this error: "One or more operations were aborted due to lack of memory", then: "The file save operation failed. Please try again, or save to a different location", after which the program crashes whatever I do.

In case anyone understands such things, the crash Event log reads:

"Fault bucket 1631980593, type 1

Event Name: APPCRASH

Response: None

Cab Id: 0

Problem signature:

P1: Vectorworks2010.exe

P2: 15.0.2.0

P3: 4b0e2827

P4: Vectorworks2010.exe

P5: 15.0.2.0

P6: 4b0e2827

P7: c0000005

P8: 001ed148

P9:

P10: "

If indeed a memory issue, I'm already on 4GB RAM, so I can only increase the page file/virtual memory, however I understand Windows may be automatically doing this anyway, and it will actually slow programs down. Anyone tried this?

Everything's in one file, 10MB, which doesn't seem excessive. It's a medium-sized medium-complexity 4 storey building. The worry is I'm still working on the shell, and haven't even started on the MEP or interior layouts, so there's a lot more information yet to go into the model. I can't issue drawings if I can't render hidden line SLVPs. It was alarming recently when I needed to issue a set of PDFs, and VW actually couldn't do this.

Thoughts:

- Turn off as many layers and classess as possible from the VPs (not really an option for Section VPs, where you're cutting through most building elements, despite limiting Depth Range to only a few meters);

- Closing all other programs to free RAM;

- [..Does VW cause a 'memory leak'..? ];

- Would unchecking "GDI + Imaging" help?

- ...?

I'll keep experimenting to try and find out more, but any thoughts are very welcome.

[Assemblage]

Further: I'm assuming splitting the file up will not help, as a SLVP will still need to process the geometry whether it's coming from the file or a referenced file(s). Is this assumption correct?

[Assemblage]

Screenshot attached of CPU and Physical memory usage when VP refuses to update.

[mike m oz]

Update to SP3 and see if that fixes the problem.

[Assemblage]

Did the following:

- Increased Page File/virtual memory

- Simplified Hole components of hybrid symbols

- Pruned off as many layers and classes as possible from every VP

- Upgraded to SP3

Not sure which one was decisive, but VPs are now updating. Have had one crash since, but not a particularly bad one.

phew..

[Assemblage]

It's mystifying however why final qual RW of the whole model should perform without incident working in design layers, but setting the SLVP to just OpenGL puts the VP update into a tailspin and the computer crashes. Seems medieval limited to hiddenline SLVPs only