Dave,

Il giorno ven 3 dic 2021 alle ore 04:16 Li Li <l...@streamnative.io.invalid>
ha scritto:

> +1 Good idea, we will take a look and do some optimizations
>
> Thanks,
> LiLi
>
> > On Dec 3, 2021, at 5:04 AM, Dave Fisher <w...@apache.org> wrote:
> >
> > Hi -
> >
> > I noticed some things about the website build.
> https://github.com/apache/pulsar/actions/workflows/ci-pulsar-website-build.yaml
> >
> > (1) It runs for nearly 2 hours.
>

why ?
is it so hard to build the website ?


> > (2) If it exceeds 2 hours then it is canceled.
> > (3) It runs every 6 hours.
> >
> > We share runners with all of the ASF.
> >
> > I think we need to change from 4 runs per day to either 2 or 3 runs.
>

+1


> >
> > We can also consider switching to the ASF’s Jenkins at
> https://builds.apache.org/job/Pulsar/ - it might give better performance.
>

We were on ASF Jenkins, but we used to use too many resources and then we
moved to GH.
I don't know the details but there are still many discussions on builds@
about the fact the Pulsar is consuming many ASF resources for CI.

We can move the WebSite stuff to Jenkins

IIUC recently the ASF started to assign dedicated Jenkins instances to
projects, we could try to go that way, and probably some company
may "donate" some resources to host the workers.

Enrico



> >
> > Regards,
> > Dave
>
>

Reply via email to