I suppose you can speed up the build by using more system packages. A "normal" build from scratch using system packages needs less than 2 hours of wall clock time, on a 5 y.o. machine with Gentoo.
On Thu, 9 Dec 2021, 05:37 julian...@fsfe.org, <julian.ru...@fsfe.org> wrote: > I rebuilt the sagemath & sagemath-dev docker images for the tags 9.3, 9.4, > and latest (alias for 9.4). I am now pushing them to the Docker Hub so they > should be available shortly. > > The build scripts seem to work still, however, the images cannot be built > on tiger lake due to https://trac.sagemath.org/ticket/32997 and each > build takes about 8h on my machine due to > https://trac.sagemath.org/ticket/32998. > > On Sunday, December 5, 2021 at 9:15:40 AM UTC-6 m.derick...@gmail.com > wrote: > >> Yeah I had to use buildx to debug why the build was failing (standard >> docker build would not give enough logs). For me doing: >> >> docker buildx create --use --name larger_log --driver-opt >> env.BUILDKIT_STEP_LOG_MAX_SIZE=50000000 docker buildx build <same commands >> I was giving to the standard docker build command> >> Just worked without any other config. I guess you could even leave out >> the "--driver-opt env.BUILDKIT_STEP_LOG_MAX_SIZE=50000000" part. >> >> Note that for me the build was failing cause of the OOM killer killed one >> of the build processes, so the only thing I needed to do to make the build >> work is assign more resources to the docker container to make it work. >> >> On Sunday, 5 December 2021 at 15:14:32 UTC+1 dim...@gmail.com wrote: >> >>> Anyone tried using buildx? >>> >>> It seems to simplify setting up the build - although on Linux it >>> appears to have extra hoops to jump through. >>> >>> On Sun, 5 Dec 2021, 14:06 Maarten Derickx, <m.derick...@gmail.com> >>> wrote: >>> >>>> Hi Julian, >>>> >>>> I didn't build a sagemath-dev image. So the first test will likely >>>> fail. I did manually test that the command line starts and that I could do >>>> computations which depend on singular pari and gap, and I also tested >>>> manually that jupyter starts and that I could do computations in it. The >>>> main purpose for building this image is that I use the sagemath docker >>>> images as base image for CI testing in a downstream package. And I needed >>>> certain features/fixes that are in 9.4 but not 9.2. And up to now all the >>>> CI tests of the downstream package pass just fine with this image. >>>> >>>> I will try to see if I can also successfully build and test a dev image. >>>> On Saturday, 4 December 2021 at 18:08:19 UTC+1 dim...@gmail.com wrote: >>>> >>>>> One way or another, it's probably a good idea to run the builder on GH >>>>> Actions - translating one yml to another should be doable. >>>>> >>>>> >>>>> On Sat, 4 Dec 2021, 16:11 julian...@fsfe.org, <julian...@fsfe.org> >>>>> wrote: >>>>> >>>>>> Hi Maarten, >>>>>> >>>>>> thanks for verifying that the Dockerfile still works. >>>>>> >>>>>> On Thursday, December 2, 2021 at 10:06:30 PM UTC-6 >>>>>> maarten...@navara.nl wrote: >>>>>> >>>>>>> In the meantime I managed to verify that aside from the gitlab CI/CD >>>>>>> there are no other things that are broken. Meaning that I managed to >>>>>>> build >>>>>>> the docker file shipped with sage just fine on my laptop. I pushed a >>>>>>> build >>>>>>> of 9.4 to https://hub.docker.com/r/mderickx/sagemath/ in case >>>>>>> anyone is interested. >>>>>>> >>>>>> >>>>>> Did you also run the tests on the docker image? There are three tests >>>>>> defined here >>>>>> https://gitlab.com/sagemath/sage/-/blob/develop/.gitlab-ci.yml#L145. >>>>>> These often failed in the past with new releases. >>>>>> >>>>>> julian >>>>>> >>>>>> -- >>>>>> >>>>> You received this message because you are subscribed to the Google >>>>>> Groups "sage-devel" group. >>>>>> To unsubscribe from this group and stop receiving emails from it, >>>>>> send an email to sage-devel+...@googlegroups.com. >>>>>> >>>>> To view this discussion on the web visit >>>>>> https://groups.google.com/d/msgid/sage-devel/6d614b77-f673-4dbf-9e38-661fc7af4247n%40googlegroups.com >>>>>> <https://groups.google.com/d/msgid/sage-devel/6d614b77-f673-4dbf-9e38-661fc7af4247n%40googlegroups.com?utm_medium=email&utm_source=footer> >>>>>> . >>>>>> >>>>> -- >>>> You received this message because you are subscribed to the Google >>>> Groups "sage-devel" group. >>>> To unsubscribe from this group and stop receiving emails from it, send >>>> an email to sage-devel+...@googlegroups.com. >>>> >>> To view this discussion on the web visit >>>> https://groups.google.com/d/msgid/sage-devel/fc867437-4b30-4efa-8d17-4f1c842bc984n%40googlegroups.com >>>> <https://groups.google.com/d/msgid/sage-devel/fc867437-4b30-4efa-8d17-4f1c842bc984n%40googlegroups.com?utm_medium=email&utm_source=footer> >>>> . >>>> >>> -- > You received this message because you are subscribed to the Google Groups > "sage-devel" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to sage-devel+unsubscr...@googlegroups.com. > To view this discussion on the web visit > https://groups.google.com/d/msgid/sage-devel/8ec620b5-78be-43c7-a58f-218877bc9426n%40googlegroups.com > <https://groups.google.com/d/msgid/sage-devel/8ec620b5-78be-43c7-a58f-218877bc9426n%40googlegroups.com?utm_medium=email&utm_source=footer> > . > -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/CAAWYfq0jANrz3RmU4TWLhQ-BeMXFJ5o0NH%2B7u94CMLPURb2krg%40mail.gmail.com.
