Cmake and rpmbuild

Cmake and rpmbuild

[Gerard Braad]

Hi All,


At the moment I am setting up CI pipelines for Ceph and ran into a
small issue; I have some memory constrained runners (2G). So, when
performing a build using do-cmake all is fine... the build might last
long, but after an hour or two I am greeted with a 'Build succeeded'
message, I gather the artifacts and all is well.

But when I do a rpmbuild, I have to rely on doing a make-dist, and
then issue a rpmbuild targeting this tarball. However, in the same
runners, this build will fail with "virtual memory exhausted". Is
there anything I can do that does not immediately involve adding more
memory?

If more information is needed, please let me know...

regards,


Gerard

-- 

   Gerard Braad | http://gbraad.nl
   [ Doing Open Source Matters ]

Re: Cmake and rpmbuild

[Kamble, Nitin A]

Gerald,
   To me this looks like a terribly crippled environment to build ceph, and I won’t bother building in such environment. IMO, It’s not worth for anybody to optimize the build process to make it work on such a crippled environment.

And to answer your question, If you can not add more memory then provide more swap space to increase the virtual memory.  You can also remove the  “-j" parameter to “make" in the spec file to avoid creating lot of processes for parallel build.

Nitin

> On Jul 28, 2016, at 6:50 PM, Gerard Braad <me@gbraad.nl> wrote:
> 
> Hi All,
> 
> 
> At the moment I am setting up CI pipelines for Ceph and ran into a
> small issue; I have some memory constrained runners (2G). So, when
> performing a build using do-cmake all is fine... the build might last
> long, but after an hour or two I am greeted with a 'Build succeeded'
> message, I gather the artifacts and all is well.
> 
> But when I do a rpmbuild, I have to rely on doing a make-dist, and
> then issue a rpmbuild targeting this tarball. However, in the same
> runners, this build will fail with "virtual memory exhausted". Is
> there anything I can do that does not immediately involve adding more
> memory?
> 
> If more information is needed, please let me know...
> 
> regards,
> 
> 
> Gerard
> 
> -- 
> 
>   Gerard Braad | http://gbraad.nl
>   [ Doing Open Source Matters ]
> --
> To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

_______________________________________________
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com

Re: Cmake and rpmbuild

[Gerard Braad]

On Fri, Jul 29, 2016 at 2:20 PM, Kamble, Nitin A
<Nitin.Kamble@teradata.com> wrote:
>    To me this looks like a terribly crippled environment to build ceph, and I won’t bother building in such environment. IMO, It’s not worth for anybody to optimize the build process to make it work on such a crippled environment.

PoC phase... but likely the conclusion is as expected. Builds run
perfectly in a 8G setup.

> And to answer your question, If you can not add more memory then provide more swap space to increase the virtual memory.  You can also remove the  “-j" parameter to “make" in the spec file to avoid creating lot of processes for parallel build.

`-j` has been altered and had no effect. Just wondering why doing a
`cmake` finishes without any problem in the same environment. And why
rpmbuild is also running something else than cmake (this is a target
for future)?

regards,
Gerard
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Cmake and rpmbuild

[Willem Jan Withagen]

On 29-7-2016 03:50, Gerard Braad wrote:
> Hi All,
> 
> 
> At the moment I am setting up CI pipelines for Ceph and ran into a
> small issue; I have some memory constrained runners (2G). So, when
> performing a build using do-cmake all is fine... the build might last
> long, but after an hour or two I am greeted with a 'Build succeeded'
> message, I gather the artifacts and all is well.
> 
> But when I do a rpmbuild, I have to rely on doing a make-dist, and
> then issue a rpmbuild targeting this tarball. However, in the same
> runners, this build will fail with "virtual memory exhausted". Is
> there anything I can do that does not immediately involve adding more
> memory?
> 
> If more information is needed, please let me know...

Hi Gerard,

Experience is that for the automake part you'd really need 8G to not run
into this problem. And be happy that your running GCC because Clang is
even more memory hungry.

--WjW