From mboxrd@z Thu Jan 1 00:00:00 1970 X-Msuck: nntp://news.gmane.io/gmane.comp.sysutils.supervision.general/2614 Path: news.gmane.org!.POSTED.blaine.gmane.org!not-for-mail From: Jeff Newsgroups: gmane.comp.sysutils.supervision.general Subject: killall test run Date: Sun, 19 May 2019 14:57:14 +0200 Message-ID: <9538981558270634@myt3-c573aa6fc782.qloud-c.yandex.net> References: <11997211556565598@myt6-27270b78ac4f.qloud-c.yandex.net> <20190501033355.6e41e707@mydesk.domain.cxm> <20190515132206.03f9736e@mydesk.domain.cxm> <20190516012214.15ffcf2e@dickeberta> <20190515210717.27b002ba@mydesk.domain.cxm> <5f5b6035-240b-e6d6-497d-d9bb945d135f@obarun.org> <5994381558140755@sas2-2074c606c35d.qloud-c.yandex.net> <57af526d-cc12-135b-218c-721d51129876@obarun.org> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit Injection-Info: blaine.gmane.org; posting-host="blaine.gmane.org:195.159.176.226"; logging-data="76370"; mail-complaints-to="usenet@blaine.gmane.org" To: super Original-X-From: supervision-return-2204-gcsg-supervision=m.gmane.org@list.skarnet.org Sun May 19 14:57:21 2019 Return-path: Envelope-to: gcsg-supervision@m.gmane.org Original-Received: from alyss.skarnet.org ([95.142.172.232]) by blaine.gmane.org with smtp (Exim 4.89) (envelope-from ) id 1hSLNQ-000JlT-57 for gcsg-supervision@m.gmane.org; Sun, 19 May 2019 14:57:20 +0200 Original-Received: (qmail 12453 invoked by uid 89); 19 May 2019 12:57:45 -0000 Mailing-List: contact supervision-help@list.skarnet.org; run by ezmlm Original-Sender: Precedence: bulk List-Post: List-Help: List-Unsubscribe: List-Subscribe: List-Id: Original-Received: (qmail 12446 invoked from network); 19 May 2019 12:57:45 -0000 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=yandex.com; s=mail; t=1558270634; bh=0prSdRWA8660XaYC8EM7pEj4KQirKHQItkm9uEMZ3LA=; h=Message-Id:Subject:In-Reply-To:Date:References:To:From; b=nXlP6VxwBAnTG8LWL/5lep7BTH4QYUKwfv2QFcMX0HEikVWJIqumbBgZIk5U+WXTZ zKmOzczaqRRQ/gXzz1QCTlnOdh/+/P/AGi0e4el5eVbjYlYTGkBmfMIQsgPjVrFpp6 kHMXNm3mpkUho8HpcUKLHQL+vwJAQEEyAAilLr0w= Authentication-Results: mxback3g.mail.yandex.net; dkim=pass header.i=@yandex.com In-Reply-To: X-Mailer: Yamail [ http://yandex.ru ] 5.0 Xref: news.gmane.org gmane.comp.sysutils.supervision.general:2614 Archived-At: 19.05.2019, 13:24, "fungal-net" : > I am glad some of you can tell more than I can about this, and since you > did I tried my weirdest of setups. This is Adélie adelielinux.org > installation on HD. Although it is confusing to me how they set this up > still, after months of following its development (beta3), there is > sysvinit on the first steps of booting then OpenRC takes over, and then > s6-supervisor handles everything running. It is like a fruit punch in > my eyes. For those that don't know this is built on musl. they use SysV init + OpenRC (with some scripts from Alpine) OpenRC is known to work with s6-svscan (this is done via the /libexec/rc/sh/s6.sh shell script). although it is better to start s6-svscan from /etc/inittab (directly or - as i prefer - via a starter shell/execline script that ensures at least the scandir exists) since SysV init can respawn processes anyway (supervised supervisor ;-). by default OpenRC (better: s6.sh) uses /run/openrc/s6-scan as the service scandir and /var/svc.d as service template dir, this can be changed easily of course since only shell scripts are involved here. when starting an s6 service it copies the service template dir into the scandir /run/openrc/s6-scan. for this to work properly an already running instance of s6-svscan (that runs on said scandir) is required. OpenRC achieves this by adding "s6-svscan" to the "need" call in the depend function of the corresponding service script. when starting s6-svscan from inittab OpenRC does not have to start it and there is no need to start the s6-svscan nor to add it to other services' dependencies. although i do not know the order in which sysvinit processes the inittab entries for a given (SysV) "runlevel". do "wait" entries precede "respawn" entries, followed by the "once" entries ? dunno, this needs a bit of hackery to make it work but this ordering/dependency problem is definitely solvable. > # kill -9 -1 on tty1 brought me back to tty1 login screen with 5 more > ttys active. So everything is respawned almost instantly to a system > just like it had just booted. Doing the same from terminal on X had the > same exact outcome. as expected. :D > Both s6/s6-rc and 66 pkgs are available through void's repositories but > s6-rc has been modified and I haven't been able to get it to work. you can find out easily about your init via $ ps -Fjlp 1 also have a look in the /proc/1 dir when procfs is mounted on /proc, the target of the exe symlink is of interest here. to see how all this is organised check with $ ps -FHjle fww BEFORE kill -1 -9 and again after sending the kill blast. > Void uses arch-like /bin /sbin --> /usr/bin, yes, i noticed this too. this is not "arch" like but was invented by fedora/ red hat instead. being the fedora testbed the lame arch distro had to follow suit immediately as is typical for them. > Adélie has more traditional 4 separate directories. this is what one would expect.