xen-devel.lists.xenproject.org archive mirror
 help / color / mirror / Atom feed
From: Wei Liu <wei.liu2@citrix.com>
To: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: Juergen Schinker <ba1020@homie.homelinux.net>,
	Wei Liu <wei.liu2@citrix.com>,
	xen-devel <xen-devel@lists.xen.org>
Subject: Re: Test Xen 4.8 RC4 not stable 27.10.16
Date: Fri, 28 Oct 2016 17:08:55 +0100	[thread overview]
Message-ID: <20161028160855.GN30231@citrix.com> (raw)
In-Reply-To: <210f9c34-2175-bd08-a345-612e25095d97@oracle.com>

On Fri, Oct 28, 2016 at 12:04:29PM -0400, Boris Ostrovsky wrote:
> On 10/28/2016 11:30 AM, Wei Liu wrote:
> > On Fri, Oct 28, 2016 at 04:01:54PM +0100, Juergen Schinker wrote:
> >>
> >> ----- On 28 Oct, 2016, at 13:07, Boris Ostrovsky boris.ostrovsky@oracle.com wrote:
> >>
> >>> I believe at least on some distros /var/run should be soft-linked to
> >>> /run, otherwise whoever cleans up those directories (the command name
> >>> escapes me right now) will only remove files from /run and leave
> >>> /var/run (and therefore /var/run/xen/xenstored.pid) untouched.
> >>>
> >>> And because xencommons checks for existence of this this file before
> >>> starting xenstored the latter never starts.
> >>>
> >>> -boris
> >> root@xen:~# ls -la /var/run/xen/xenstored.pid 
> >> -rw-r----- 1 root root 6 Oct 28 15:15 /var/run/xen/xenstored.pid
> >>
> >>   
> >>    33 ?        00:00:00 xenwatch
> >>    34 ?        00:00:00 xenbus
> >>    45 ?        00:00:00 xenbus_frontend
> >>   785 ?        00:00:00 xen_pciback_wor
> >>  1137 ?        00:00:00 xenwatchdogd
> >>  1169 ?        00:00:00 xen-init-dom0
> >>  1175 ?        00:00:00 xenconsoled

And this ^ 

xenconsoled requires xenstored.service, so if xenstored is not available
xenconsoled wouldn't be started, I think.

xen-init-dom0 as well.

> >>
> >> Thats exactly what I'm sayin , maybe we need a more intelligent check to see if xenstored is running
> >>
> >> not just a simple pid check
> >>
> > But ... the pid check is the usual way of checking if a daemon is
> > active.
> >
> > I think a bit work is required to work out why xenstored.pid stays
> > across reboot -- it is not supposed to work like that on a FHS compliant
> > system.
> 
> I remember now: it's systemd-tmpfiles that creates and cleans up those
> files. I think the way it works is it checks
> /usr/lib/tmpfiles.d/var.conf for the rules. On my fedora24:
> 
> L /var/run - - - - ../run
> 
> Which IIUIC means that /var/run should be linked to ../run (i.e. /run)
> during boot. Or something along these lines. Perhaps because when we
> shut down xenstored.pid is still there and so the link cannot be made.

It's the same on my Debian Jessie box, but it works just fine, FWIW.

Wei.

> 
> 
> -boris
> 
> >
> > And how do other daemons work on your test host, presumably they will
> > see stale pid files as well.  Are there any other pid files under
> > /var/run? If so, do the corresponding daemon run properly?
> >
> > Wei.
> >
> >> J
> 
> 

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

  reply	other threads:[~2016-10-28 16:08 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-10-26 18:47 Test Xen 4.8 RC4 SUCCESS 26.10.16 Juergen Schinker
2016-10-27  9:53 ` Wei Liu
2016-10-27 22:01   ` Test Xen 4.8 RC4 not stable 27.10.16 Juergen Schinker
2016-10-28  9:29     ` Wei Liu
2016-10-28 13:07       ` Boris Ostrovsky
2016-10-28 15:01         ` Juergen Schinker
2016-10-28 15:30           ` Wei Liu
2016-10-28 16:04             ` Boris Ostrovsky
2016-10-28 16:08               ` Wei Liu [this message]
2016-10-27 13:58 ` Test Xen 4.8 RC4 SUCCESS 26.10.16 Dario Faggioli

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20161028160855.GN30231@citrix.com \
    --to=wei.liu2@citrix.com \
    --cc=ba1020@homie.homelinux.net \
    --cc=boris.ostrovsky@oracle.com \
    --cc=xen-devel@lists.xen.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).