Re: [Scilab-users] Increasing Scilab default Java heap memory to 512 MB?
Le 20/12/2020 à 18:53, Antoine Monmayrant a écrit :
On 19/12/2020 16:24, Samuel Gougeon wrote:
Le 16/12/2020 à 14:16, Antoine Monmayrant a écrit :
That's the very first thing we change after installing scilab on all
the computers in our group, so I agree we need to change it.
I conclude that 512 MB is enough. Otherwise you would not have
experienced and reported the issue.
Wait, what?
I don't see how you conclude 512 is enough?
You are right: 512 MB is enough to pass the example reformulated by
Stéphane and Christophe, not to pass its original version with
driver("PNG").
But when i increase the java heap to 1 GB, your original script passes
without error:
--> exec('test.sce', -1)
"SCREEN: Time to plot = 1.2227456s"
"HEADLESS: Time to plot = 33.772641s"
By the way, beside the driver() discussion, the Java memory consumption
of the grayplot family is known for a long time:
http://bugzilla.scilab.org/8747 http://bugzilla.scilab.org/13150 and so on.
___
users mailing list
users@lists.scilab.org
http://lists.scilab.org/mailman/listinfo/users
Re: [Scilab-users] Increasing Scilab default Java heap memory to 512 MB?
On 19/12/2020 16:24, Samuel Gougeon wrote: Le 16/12/2020 à 14:16, Antoine Monmayrant a écrit : That's the very first thing we change after installing scilab on all the computers in our group, so I agree we need to change it. I conclude that 512 MB is enough. Otherwise you would not have experienced and reported the issue. Wait, what? I don't see how you conclude 512 is enough? In my case, I set it to 2GB: A patch with this value is submitted for Scilab 6.1.1 @ https://codereview.scilab.org/21650 Cool! Antoine Samuel ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users
Re: [Scilab-users] Increasing Scilab default Java heap memory to 512 MB?
Le 16/12/2020 à 14:16, Antoine Monmayrant a écrit : That's the very first thing we change after installing scilab on all the computers in our group, so I agree we need to change it. I conclude that 512 MB is enough. Otherwise you would not have experienced and reported the issue. A patch with this value is submitted for Scilab 6.1.1 @ https://codereview.scilab.org/21650 Samuel ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users
Re: [Scilab-users] Increasing Scilab default Java heap memory to 512 MB?
That's the very first thing we change after installing scilab on all the computers in our group, so I agree we need to change it. Is there a way to make people aware of this settings and tell them that it's related to plotting? Like displaying a tip saying that they can crank it up when they need to display/export large plots and that it will add up to the memory used by scilab? Antoine On 15/12/2020 19:09, Samuel Gougeon wrote: Opening a new thread after http://mailinglists.scilab.org/Scilab-users-Headless-graphics-in-scilab-time-memory-issues-tp4041133p4041160.html Le 15/12/2020 à 18:58, Samuel Gougeon a écrit : Le 03/12/2020 à 09:59, Antoine Monmayrant a écrit : Hello again, Just to add to my previous message: increasing nx or ny in my script can crash scilab during the call to xend(). On my small local machine, this occurs for: nx=4000; ny=3000; for which scilab takes ~7 seconds to create the png in "normal" mode... Can any of you confirm this bug? The crash is due to a too small Java Heap of memory. When increasing it through the user's Preferences from 256 MB to 1 GB, the PNG export works as fast as expected. By the way, everyone, what do you think about increasing the default Scilab java Heap from 256 MB to 512 MB? This default 256 MB was set a long time ago, when on computers the average RAM what much smaller than today... Samuel ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users
[Scilab-users] Increasing Scilab default Java heap memory to 512 MB?
Opening a new thread after http://mailinglists.scilab.org/Scilab-users-Headless-graphics-in-scilab-time-memory-issues-tp4041133p4041160.html Le 15/12/2020 à 18:58, Samuel Gougeon a écrit : Le 03/12/2020 à 09:59, Antoine Monmayrant a écrit : Hello again, Just to add to my previous message: increasing nx or ny in my script can crash scilab during the call to xend(). On my small local machine, this occurs for: nx=4000; ny=3000; for which scilab takes ~7 seconds to create the png in "normal" mode... Can any of you confirm this bug? The crash is due to a too small Java Heap of memory. When increasing it through the user's Preferences from 256 MB to 1 GB, the PNG export works as fast as expected. By the way, everyone, what do you think about increasing the default Scilab java Heap from 256 MB to 512 MB? This default 256 MB was set a long time ago, when on computers the average RAM what much smaller than today... Samuel ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users
