Export Project Failed (Error 100013)

[Anonymous]

I keep getting this problem where I am exporting a registered .rcp file to .rcs. My rcp file itself is 129GB in size. The point cloud itself was registered with Faro Scene 5.5.0.44203. It was exported from Faro Scene as a .rcp file. Now what I want to do is to create .rcs file. My Recap version is Recap 360 Ultimate 2.2.0.25. Last time I checked my screen, it has reached about 40% completion. I left it overnight and by the time I checked back in the morning, I saw this Error 100013 message. I didn't see Autodesk Recap 360 Error Report page and my Recap didn't crash. 

Below, I attached the screen shot of the error message I got.

Computer specs: Core i7-4900MQ @ 2.8GHz, 32GB RAM, AMD FirePro M6100 FireGL V, Windows 7 

Thanks in advance.

[ryan.frenz]

This appears to be memory-related, although that's surprising considering you have 32GB.  Did you notice the RAM utilization at the time of the error message?

I've forwarded this to the engineering team in charge of the related code.  I'll update this thread if/when we know more.

Ryan

[Anonymous]

At the exact time of the error message pop out? I don't know that since it happened during the night after I left my office. But at the I left, at which it was 40% completion, I remembered my memory usage was about low 20 GB, probably around 21-22GB. I didn't have other programs open at all, just Recap.

I started the rcs export process at 9AM and I left at 5.30PM at which it reached about 40% completion. Even then, for a 129 GB of rcp file and about 8.5 hours for which my computer run only Recap. Does it seem normal for Recap to only reach 40% completion after 8.5 hours?

[Anonymous]

One more thing to add.

I am able to export the same rcp file to pcg just fine.This just happen if I export my rcp to rcs. 

[CharlesBerteaux]

I get this too. only exporting RCS. I have dumped all my temp files, local revit files, everyhting i possibly can, reboot, and try again with no luck.

Pretty much the same setup in laptop as the guy above!

[vidanom]

Hi,

Sorry to hear that you are having trouble when exporting scan files.

Recently there was a new hotfix released that mentions that some of the RCS export errors were fixed, and maybe it will work for your case too:

http://forums.autodesk.com/t5/recap-general-discussion/recap-2-2-1-1-update-release-notes/td-p/60437...

It might worth to try.

Thank you,

Mitko

[Anonymous]

I'm also still experiencing this issue in 3.0.2.12, it seems 1 in 4 exports fail with the same error message. Very frustrating. Does the maximum memory in settings need to be raised for a 40GB file with 80scans? I've edited every scan in the project removing small and large sections depending how the scans were interacting. Does Recap have a maximum limit of editing, as in good for light stuff but don't expect heavy edits. I have made sure to Save As + Remove points before trying to export. I've tried putting the temporary file location in both the operating and working drives. Our machine has 64GB of RAM and 200GB free in the working drive, about the same in the C... that should be enough to get the job done..otherwise how is anyone getting anything out at all reliably? 

I have created a resource report for the duration of the export using Windows tool Performance Monitor if it is at all helpful to the dev team or to try and figure out if it's our computer specifically. Just need a place to send it to if applicable. 

Thanks,

[Anonymous]

Hi everyone,

I have pretty much exact same issue with the original poster. I got this "Export Project Failed, error 100013" message when I was exporting RCS file.

It surprised and frustrated me because Autodesk gave a hotfix for version 2.2.1.1 but the exact same problem happen in the newer version. I have Recap 360 Version 3.0.2.12.

To review the project file I worked on. All 135 scans were registered using Faro Scene 5.5.3.16 and were exported from Scene as PTX files. Then I created new project in Recap 360 Ver. 3.0.2.12 by importing those PTX files. Because of what I read on this forum about how slow Recap is after several point cleanups, I immediately exported RCS once the PTX files were indexed. The RCS export took a while but still successfully exported. Then I cleaned up the unnecessary points and created save copy as with the option to remove all deleted points to compact the file. From there, I tried to export the RCS file again but I got the 100013 error message. It did however created temp folder with lots of NODEPTS files (Please refer to the screenshot below).

Is this a recurring bug?

It didn't create error report, the error message just popped up.

Desktop spec: Xeon E5-2687W 3.40Ghz, 128GB RAM, Windows 7 Enterprise, GeForce GTX Titan Black 4095MB

[vidanom]

Hi, thank you for reporting this.

Our team was able to reproduce it and the same was reported.

For speeding up the export process there have been some improvements that are planned to go out in the next upcoming release.

Thank you,

Mitko

[Anonymous]

Any estimate on time frame? I was able to export two separate files earlier this morning with very light editing in about 2.5 hours each for 65 and 85 scans. Should we be looking at not heavily cleaning in Recap for the time being? Odd that this just started to become an issue.. Also the workaround with exporting to RCS only solves half the battle  because once the scans are unified you lose the ability to individually isolate bad spray and lighting. 

-Kyle

[ryan.frenz]

Hey Kyle,

I can confirm that we have changes on both fronts upcoming:

• Enhancements to unify pipeline to use available RAM instead of temp files
• Fix for the intermittent freeze/crash when exporting to an external/slower drive.

(edit: you asked about timeframe - I'm not allowed to talk dates, but we're trying to get this out as soon as possible)

Note, though, that your export performance is also tied heavily to the complexity of the edit stack.  Export basically loads every point from disk, passes it through each edit/clip/hide operation, and indexes a new file with the points that pass.  To improve that, you could optimize the project more frequently - for example, clean one or two scans, then optimize, repeat.  The optimization pass clears the delete stack, which is a big part of the performance considerations mentioned above.  Points deleted via the automatic cleanup service are an extreme case of this (as the resulting delete operation is quite complex).

Also - I will triple-check, but none of the above should be new since almost a year.  If you're feeling like performance is much worse lately, it's likely due to what I describe above.

-Ryan

[Anonymous]

Thanks for the response Ryan.

First point you made, do you see the ability for Recap to effectively utilize 128GB of RAM on larger projects? Ie. would it speed things up and necessitate less saving? With the second point you made, just as a head up for end user data we are running a 950 Pro m.2 in the C: Drive and an 850 Evo in the working drive. But both points are great to hear.

I will go ahead and run that today and post back with detailed results. 

-Kyle

[ryan.frenz]

Yes, currently the unifier uses temp files based on static limits for how much to keep in memory.  The changes will make that mechanism be dynamic based on available RAM.  We have implemented this for the unification step (the 'nodepts' files) - should be a pretty nice improvement, especially for large projects on high-RAM machines.

For the bug, what we found was a race condition when setting up the export - it can happen on any drive/speed but was more obvious the slower the drive (and it probably has something to do with how the OS handles multiple drives).  It will be fixed regardless.

-Ryan

[CharlesBerteaux]

Is this going to be an issue with every single version moving forward??? I am currently on version 3.1.2.1. and this is still an issue, and could not be happening at a worse time!!!!!!

[ashwini.murthy]

Hello,

Are you seeing "Error:100013"? That usually implies there is not enough disk space. 

Unified export writes a lot of temp files and fills up a lot of space very quickly. On huge projects, I recommend having enough disk space before you start the export (at least 3 times size of the project). Once the export is complete, the temp files will be cleaned up. 

I will put in a request to make the error message more meaningful :).

Hope that helps.

Ashwini

[jared.white]

I just ran into this last night as well.  I have 128 GB of Ram and 2TB free hard drive space.  Shouldn't be an issue with space.  I am trying to export about 50 GB to a single RCS, and it failed at 58%. I am running Recap 3.1.3.4

[ashwini.murthy]

Hi Jared,

Thanks for reporting. We will try to recreate this in-house.

Ashwini

[Anonymous]

Same problem. It definitely has to do with the size of the region I'm trying to export. Thanks.

ReCap Version 3.1.3.4

Working on a 931GB drive

Processor Intel(R) Core(TM) i7-6700HQ CPU @ 2.60GHz, 2592 Mhz, 4 Core(s), 8 Logical Processor(s)

Installed Physical Memory (RAM) 32.0 GB

[ashwini.murthy]

Hello,

What is the size of your project? We are aware and investigating a similar issue where you can run out of memory during exporting huge projects, but its rare.

I have also seen this happen when you run out of disk space. Ensure you have enough space (more than 2-3 times size of the project).

I noticed from another thread you mentioned using ReCap 3.1. The latest version as of today is 4.0.0.28. It has a bunch of bug fixes, I strongly suggest upgrading to it. 

Ashwini

[Patrick_Aps_9121]

Also export on an Internal SSD, if possible.

I have had errors and long waiting time when exporting to an USB 3.0 external HD, even a crash and loss of the USB HD.

Version 4.0.n seems much faster on exporting Regions to unified RCS files.