From mboxrd@z Thu Jan 1 00:00:00 1970 From: YOANN P Subject: RE: Questions regarding offloading ( unprivileged setup , parallel builds ) Date: Mon, 5 Feb 2018 16:13:31 +0000 Message-ID: References: , <87eflzhfjo.fsf@gnu.org> Mime-Version: 1.0 Content-Type: multipart/alternative; boundary="_000_DB6P18901MB00221A60916A2DCFBA85D80EDBFE0DB6P18901MB0022_" Return-path: Received: from eggs.gnu.org ([2001:4830:134:3::10]:42505) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1eijOm-0004Iq-9z for guix-devel@gnu.org; Mon, 05 Feb 2018 11:13:41 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1eijOg-0002Lm-CS for guix-devel@gnu.org; Mon, 05 Feb 2018 11:13:40 -0500 Received: from mail-oln040092068020.outbound.protection.outlook.com ([40.92.68.20]:45175 helo=EUR02-HE1-obe.outbound.protection.outlook.com) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1eijOf-0002Dm-UC for guix-devel@gnu.org; Mon, 05 Feb 2018 11:13:34 -0500 In-Reply-To: <87eflzhfjo.fsf@gnu.org> Content-Language: fr-FR List-Id: "Development of GNU Guix and the GNU System distribution." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: guix-devel-bounces+gcggd-guix-devel=m.gmane.org@gnu.org Sender: "Guix-devel" To: "guix-devel@gnu.org" --_000_DB6P18901MB00221A60916A2DCFBA85D80EDBFE0DB6P18901MB0022_ Content-Type: text/plain; charset="Windows-1252" Content-Transfer-Encoding: quoted-printable Hi, Thanks you for yours answers :-) > > Indeed, though if you=92re customizing things anyway, you could run: > > ./configure --prefix=3D/my/prefix --sysconfdir=3D/my/etc > Thanks, I missed that part sorry > > Several machines can be used at the same time, no worries. :-) > It is really nice if each dependencies could be build among different machi= nes at the same time :-) Anothers questions regarding the way the offload work: - does the machines.scm is read at the start of the daemon or read each tim= e the hook is called ? (just to be sure because the sources let me think it= is read each time and is what i want) - if a machine disappear during a build, does that build is retry on anothe= r machine ? is there a retry parameter ? - Is there any project to had a parametrable post/pre offload-hook ? (Could= be used to start preemptive cloud instances before a build and fill the "m= achines.scm" and shutdown instances at the end) - No problem to use multiple daemon with the same store ? (I'm not sure of = it because I never seen this kind of implementation on web, could be very u= seful if it not yet possible) We could imagine a S3 storage for the store / state directories and when we= need compute resources to build packages, only start preemptible instances= who mount the same store through S3QL. This way we only pay the storage for state / store directories Best regards, Yoann --_000_DB6P18901MB00221A60916A2DCFBA85D80EDBFE0DB6P18901MB0022_ Content-Type: text/html; charset="Windows-1252" Content-Transfer-Encoding: quoted-printable

=

Hi,

Thanks you for yours answers :-)

>

> Indeed, though if you=92re customizing things anyway, you could r= un:

>

> ./configure --prefix=3D/my/prefix --sysconfdir=3D/my/= etc

>

Thanks, I missed that part sorry

>

> Several machines can be used at the same time, no worries. = :-)

>

It is really nice if each dependencies could be build among diff= erent machines at the same time :-)

Anothers questions regarding the way the offload work:

- does the machines.scm is read at the start of the daemon or read eac= h time the hook is called ? (just to be sure because the sources let me thi= nk it is read each time and is what i want)

- if a machine disappear during a build, does that build is retry on a= nother machine ? is there a retry parameter ?

- Is there any project to had a parametrable post/pre offload-hook ? (= Could be used to start preemptive cloud instances before a build = and fill the "machines.scm" and shutdown instances at the end)

- No problem to use multiple daemon with the same store ? (I'm not sur= e of it because I never seen this kind of implementation on web, could be v= ery useful if it not yet possible)

We could imagine a S3 storage for the store / state directories and wh= en we need compute resources to build packages, only start preemptible inst= ances who mount the same store through S3QL.

This way we only pay the storage for state / store directories

Best regards,

Yoann

--_000_DB6P18901MB00221A60916A2DCFBA85D80EDBFE0DB6P18901MB0022_--