From mboxrd@z Thu Jan 1 00:00:00 1970 X-Msuck: nntp://news.gmane.io/gmane.comp.sysutils.supervision.general/585 Path: main.gmane.org!not-for-mail From: Laurent Bercot Newsgroups: gmane.comp.sysutils.supervision.general Subject: Re: Logging everything in the database Date: Thu, 23 Sep 2004 12:36:26 +0200 Message-ID: <20040923103626.GA23518@skarnet.org> References: <1095929269.41528db5a0c17@mail.games-expert.com> NNTP-Posting-Host: deer.gmane.org Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Trace: sea.gmane.org 1095935778 19562 80.91.229.6 (23 Sep 2004 10:36:18 GMT) X-Complaints-To: usenet@sea.gmane.org NNTP-Posting-Date: Thu, 23 Sep 2004 10:36:18 +0000 (UTC) Original-X-From: supervision-return-824-gcsg-supervision=m.gmane.org@list.skarnet.org Thu Sep 23 12:36:06 2004 Return-path: Original-Received: from antah.skarnet.org ([212.85.147.14] ident=qmailr) by deer.gmane.org with smtp (Exim 3.35 #1 (Debian)) id 1CAQxK-00086Y-00 for ; Thu, 23 Sep 2004 12:36:06 +0200 Original-Received: (qmail 23625 invoked by uid 76); 23 Sep 2004 10:36:26 -0000 Mailing-List: contact supervision-help@list.skarnet.org; run by ezmlm List-Post: List-Help: List-Unsubscribe: List-Subscribe: List-Archive: Original-Received: (qmail 23619 invoked by uid 1000); 23 Sep 2004 10:36:26 -0000 Mail-Followup-To: supervision@list.skarnet.org Original-To: supervision@list.skarnet.org Content-Disposition: inline In-Reply-To: <1095929269.41528db5a0c17@mail.games-expert.com> User-Agent: Mutt/1.4i Xref: main.gmane.org gmane.comp.sysutils.supervision.general:585 X-Report-Spam: http://spam.gmane.org/gmane.comp.sysutils.supervision.general:585 > I wanted to have it launched at the system boot, but I still wanted it to be > supervised. So I fell back on using a fifo and opening it RDWR on Linux to no > block the main process (writing to a pipe not opened RDWR when the reader isn't > started blocks...). > So tell me about this idea, what do you think about it, and what could be wrong > with it... I've done the exact same thing, to log svscan's stderr using a supervised multilog. Except that O_RDWR on a fifo isn't portable - it has no real Unix meaning; you'd better use 2 fds, one reading, one writing. I've designed a small tool to do just that: see -- Ska