Large builds with poudriere

[Chris]

On 2021-05-20 05:47, Fernando Apesteguía wrote:
> El jue., 20 may. 2021 14:45, Alexis Praga <alexis.praga at gmail.com> escribió:
> 
>> 
>> Hi guys,
>> 
>> I've recently been using poudriere to test news ports but this resulted in
>> building two large ports, namely:
>> - devel/llvm [1]
>> - lang/rust [2]
>> 
>> Now, lang/rust takes 2 hourse to compile. So testing for 13.0, 12.2 and
>> 11.4 means around 6 hours in total, a lot of time for my desktop
>> computer. Especially if the dependencies are upgraded later on.
>> 
>> Do you have any tips to deal with such long build time ? Apart from
>> have a dedicated server :)
What? You don't already have a dedicated build server? ;-)
>> 
> 
> Patience :-)
> 
> Package seeding in poudriere is a long awaited feature.
So where is the
pkg install llvm && rust
before starting the build job option?

In all honesty. This is my sole reason for not choosing
poudriere. Unless I'm performing MAINTAINER work on my ports.
Even with 8 cores, 16 threads @4.2Ghz && 128Gb RAM. It's a
daunting task. I generally use jail(8) otherwise.

--Chris
> 
> 
>> Thanks,
>> 
>> 
>> [1] For https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=233010
>> [2] For https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=255233
>> 
>> --
>> Alexis Praga
>> _______________________________________________
>> freebsd-ports at freebsd.org mailing list
>> https://lists.freebsd.org/mailman/listinfo/freebsd-ports
>> To unsubscribe, send any mail to "freebsd-ports-unsubscribe at freebsd.org"
>> 
> _______________________________________________
> freebsd-ports at freebsd.org mailing list
> https://lists.freebsd.org/mailman/listinfo/freebsd-ports
> To unsubscribe, send any mail to "freebsd-ports-unsubscribe at freebsd.org"
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 488 bytes
Desc: OpenPGP digital signature
URL: <http://lists.freebsd.org/pipermail/freebsd-ports/attachments/20210520/6d8a6ad7/attachment.sig>