Announcement

**georgi.zhekov** · 13-04-2023, 06:29 AM

Hmm, besides the error you got, there is a warning about open geometry, which might cause some issues:

[WARNING] The following objects are not closed meshes! This may cause unexpected simulation results:

2023-04-12 18:11:44: 0: STDOUT: D2@node_2

2023-04-12 18:11:44: 0: STDOUT: E@node_12

In order for the sim to work properly you need proper geometry (no single or overlapping faces, no open edges, etc.) otherwise strange things can happen. Does it work properly if you fix the geometry?

**Joelaff** · 13-04-2023, 07:19 AM

It works when using max instead of standalone to do the sim.

The bad geo is made by using a push modifier (in) too far that the geo overlaps.

I’ll try standalone without the push if I get a chance. Too many deadlines making trouble shooting difficult

**Svetlin.Nikolov** · 19-04-2023, 10:43 PM

Ah, does the geometry with the push modifier exist in the scene you sent to Support? I tried a simple setup with just a sphere and a Push modifier, but couldn't break it, so there must be something more specific...

**Joelaff** · 19-04-2023, 11:06 PM

I am pretty certain the push modifier is in that scene... and it is enabled, and when I try to render that particular scene as is I get the problems. I pulled the scene from the Deadline repository.

Note the geometry is trashed by the push modifer... But it rendered OK in Max, and it simulated like I wanted it. So I wasn't concerned about it looking messed up (as the modifier was off for other beauty renders, etc.)

**Svetlin.Nikolov** · 19-04-2023, 11:35 PM

Ah, but if it does something in the standalone simulation, then it's not the geometry that shows the
SimBodyGeomStnd::build no geometry
error (I have to add to the log the name of the node that throws this error). If it has no geometry, then it would not participate in the simulation at all.

**Joelaff** · 19-04-2023, 11:55 PM

Not sure I understand.

It would sim OK in Max, whether max interactive, or Max via Deadline. Then I got those errors when I tried to simulate using standalone through deadline. And the simulation terminated when I tried to simulate on the local machine using standalone (via the Phoenix menu), though I couldn't read the error because the cmd window closed and I did not (yet) know where the logfile was.

Oh, and yes, any more info you can put in the logfile is always a good thing. Node names, etc. because that did help me track down that disabling the push worked fine (but I would have guessed that anyway, since it was mangled geo (intersecting polys).

**Svetlin.Nikolov** · 20-04-2023, 03:06 AM

Huh, interesting why disabling the push modifier fixed it. My idea was that if it writes in the log file that there was no geometry, then the node with no geometry shouldn't even effect the standalone simulation - it would be as if it's erased from the scene.

Will try to reproduce it with the scene from Support...

**Joelaff** · 20-04-2023, 07:50 AM

Sounds good. I will see if I can retry with the new nightly we installed from apr 18.

Announcement

Weird Error with Standalone Phoenix and Deadline

Weird Error with Standalone Phoenix and Deadline

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment