Vray 2.0: Excessive RAM usage

[RSI]

Any news about the patch?
We finish our job, so at the end, I had to shut down the displacement, and lowered the size of the textures.
We notice that the big problem was for the lightcache, we have 12 gigas of memory on each workstation and we are used to put the memry limit to 10, so after a restart of the computer, first lightcache go to 596h, after 2 min stop it, relaunch wait 2 min (it says 5 or 6 hours) stop it, relaunch and it says 30 min. Look like a car engine.

[BBB3]

you should email vlado about the patch.

[pixela]

Bertrand, did you receive the patch? I still didnt.

[scheinbarreal]

same problem here.

we use distributed render to render the finale Images.
every time a computer runs out of memory we clear the memory by restarting the vrayspanner... it is dirty but it helps.

[RSI]

Hello,
Vlado, I send you an email for the patch, you might be busy right now, but is it posible to have it please?

[vlado]

When did you send it?

Best regards,
Vlado

[RSI]

13/06, 27/06 , I read somewhere that the next build were for autonm, so if it's possible to have a new build before, it's nice.
Thanks

[vlado]

You should have mail now.

Best regards,
Vlado

[RSI]

Thanks!
I'm downloading it, good we

[chipnuts]

Same problem here after VRay 2.1 switch
For this project, 24 GB ram is not enough.
VRay start with, I don't know, 19 GB of RAM and after some hours... HHD swapping

I use obviously QMC+QMC with little buckets size, low max tree and high than default Face/level, but I not solve the problem...

It's interesting see with the "old" RAM method the difference...

Bye...

We are running into the same problem with the same settings. Render farm all with 24GB ram. Their page files maintain at 24GB at first, but after they render 8 frames the page file starts to grow eventually hitting the ceiling of 75GB, then they all freeze up and have to be completely restarted. Some become completely unresponsive and have to be shut off via power button.

Any chance there's an update to this? I saw in the changelog for 2.20.10 that there were memory/multithreading bug fixes and I'm not sure if this was added. We can essentially only render 15-20 frames per node before they reach their pagefile limit. We could raise the pagefile limit but some of of the machines crash before that because their drives only have 40GB of free space. This is on 2.20.10 and it starts adding to the pagefile after 8 frames and during the Building Static Raycast Accelerator phase. The previous frames before that are no problem. Even if we render different sections of the animation, it takes 8 frames, then the pagefile starts filling up. At least with 2.20.10, it's releasing the pagefile ram after we suspend the backburner job.

Vlado, can I email you and get the fix or has this been implemented in 2.20.10 and something else may be wrong?

EDIT!: Sorry, I meant 2.20.03, NOT 2.20.10.

[vlado]

Yes, please email me and we'll see what can be done about it.

Best regards,
Vlado

[chipnuts]

Thanks, Vlado! Email is sent. Again, if can't get to it this weekend, no biggie since we're not in 'crunch mode' on this one yet. Just trying to work the kinks out before we try more drastic options. If it's a quick fix to send on, that'd be great as I'm babysitting the queue this weekend to get out what we can over the weekend.

Thanks agin!

[vlado]

Like I wrote in the email, your issue seems to be somewhat different; still, will see what can be done...

Best regards,
Vlado

[chipnuts]

Thanks for the help, Vlado! Much appreciated. I'll follow up if we discover something. Do you want me to continue to post here, start a new thread, or just continue via email?

[vlado]

At this point email is better...

Best regards,
Vlado