[lxc-users] OpenVSwitch compiling on CentOS 6.8

Fajar A. Nugraha list at fajar.net
Thu Sep 8 11:24:18 UTC 2016


On Thu, Sep 8, 2016 at 5:50 PM, Ruzsinszky Attila <
ruzsinszky.attila at gmail.com> wrote:

> Hi,
>
> I don't know what could be the problem.
>
> Host is Ubuntu 16.04 64 bit
> ii  lxc
> 2.0.4-0ubuntu1~ubuntu16.04.2
> rattila at fcubi:~$ uname -a
> Linux fcubi 4.4.0-34-generic #53-Ubuntu SMP Wed Jul 27 16:06:39 UTC 2016
> x86_64 x86_64 x86_64 GNU/Linux
>
> Container:
> [rattila at cos64 ~]$ uname -a
> Linux cos64 4.4.0-34-generic #53-Ubuntu SMP Wed Jul 27 16:06:39 UTC 2016
> x86_64 x86_64 x86_64 GNU/Linux
>
> I wanted to build OVS 2.5.0 rpm:
>
> [root at cos64 openvswitch-2.5.0]# rpmbuild -bb -D 2.6.32-642
> rhel/openvswitch.spec
> ...
> 1725: ovn-sbctl - test                                ok
> 1726: ovn-controller - ovn-bridge-mappings            ok
> 1727: ovn-controller-vtep - test chassis              skipped (
> ovn-controller-vtep.at:113)
> 1728: ovn-controller-vtep - test binding 1            skipped (
> ovn-controller-vtep.at:180)
> 1729: ovn-controller-vtep - test binding 2            skipped (
> ovn-controller-vtep.at:244)
> 1730: ovn-controller-vtep - test vtep-lswitch         skipped (
> ovn-controller-vtep.at:283)
> 1731: ovn-controller-vtep - test vtep-macs 1          skipped (
> ovn-controller-vtep.at:335)
> 1732: ovn-controller-vtep - test vtep-macs 2          skipped (
> ovn-controller-vtep.at:406)
>
> Waiting for something since some days. :-(
>
> 2700 pts/4    S      0:00 -bash
> 14824 pts/4    S+     0:00 rpmbuild -bb -D 2.6.32-642 rhel/openvswitch.spec
> 16544 pts/4    T      0:00 /bin/sh ./tests/testsuite -C tests
> AUTOTEST_PATH=util
> 16549 pts/4    T      0:00 /bin/sh ./tests/testsuite -C tests
> AUTOTEST_PATH=util
> 16550 pts/4    T      0:00 /bin/sh ./tests/testsuite -C tests
> AUTOTEST_PATH=util
> 16551 pts/4    T      0:00 cat
> 16553 pts/4    T      0:00 /bin/sh ./tests/testsuite -C tests
> AUTOTEST_PATH=util
> 16556 pts/4    T      0:00 /usr/bin/perl
> 21876 ?        Ss     0:00 sshd: rattila [priv]
> 21878 ?        S      0:00 sshd: rattila at pts/5
> 21879 pts/5    Ss     0:00 -bash
> 21895 pts/5    R+     0:00 ps ax
> 25797 pts/4    S+     0:00 /bin/sh -e /var/tmp/rpm-tmp.0qU6e1
> 25798 pts/4    S+     0:00 make check TESTSUITEFLAGS=-j2
> 25799 pts/4    S+     0:00 make check-recursive
> 25800 pts/4    S+     0:00 /bin/sh -c fail=; \?if (target_option=k; case
> ${targe
> 26420 pts/4    S+     0:00 make check-am
> 26522 pts/4    S+     0:00 make check-local
> 26523 pts/4    S+     0:00 /bin/sh ./tests/testsuite -C tests
> AUTOTEST_PATH=util
>
> I don't know it is an LXC problem or not.
>
>

Without knowing what ./tests/testsuite actually do, it's hard to tell. It's
possibly that the code is doing something not available in lxc/lxd (e.g.
loading kernel module), and doesn't handle failure correctly (e.g. keep
waiting instead of saying fail).

A workaround is to edit your SRPM (or possible the makefile, by creating an
additional patch), and comment-out the test code (there should be "make
check" or "make test" somehwere)

-- 
Fajar
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linuxcontainers.org/pipermail/lxc-users/attachments/20160908/f4cf7141/attachment.html>


More information about the lxc-users mailing list