On Sun, 2021-09-12 at 07:55 -0700, akuster808 wrote:
> 
> On 9/12/21 4:46 AM, Richard Purdie wrote:
> > We made the autobuilder sstate public a while ago but it wasn't useful 
> > after we
> > switched to hash equivalence by default. I'm pleased to be able to report 
> > that
> > we now have a read-only public server for the hash equivalence. We did have 
> > to
> > fix a small API omission and a couple of sstate related bugs but I'm 
> > pleased to
> > be able to report it is now operational and usable. An example config would 
> > be:
> > 
> > SSTATE_MIRRORS ?= "file://.* 
> > http://sstate.yoctoproject.org/dev/PATH;downloadfilename=PATH";
> > BB_HASHSERVE_UPSTREAM = "typhoon.yocto.io:8687"
> > 
> > I've included an example image build below where it took 11 minutes for a 
> > core-image-minimal image build. Obviously this is only of benefit if you're
> > building configurations that match the project autobuilder although hash
> > equivalence could be useful alone in reducing some build times.
> 
> What branches that support hash equivalence will be available other than
> Master?

The question doesn't make sense as hash equivalence doesn't have a "branch".
We're sharing any hash equivalence data generated from builds on the
autobuilder.

> > I have updated the local.conf sample file in poky to match the new release.
> > We're not widely highlighting that information as yet as it remains to be 
> > seen
> > how well it scales and what the hosting impacts for it will look like but 
> > it is
> > there to experiment with to a wider audience such as this list.
> 
> The should improve the out-of-the-box experience.

Ultimately that is the hope.

Cheers,

Richard

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#155953): 
https://lists.openembedded.org/g/openembedded-core/message/155953
Mute This Topic: https://lists.openembedded.org/mt/85549949/21656
Group Owner: openembedded-core+ow...@lists.openembedded.org
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub 
[arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Reply via email to