From: Ian Campbell <ian.campbell@citrix.com>
To: ian.jackson@eu.citrix.com, xen-devel@lists.xen.org
Subject: Re: [PATCH OSSTEST v4] ms-flights-summary: Produce an HTML report of all active flights
Date: Fri, 18 Sep 2015 12:14:25 +0100 [thread overview]
Message-ID: <1442574865.18856.290.camel@citrix.com> (raw)
In-Reply-To: <1442572034.18856.287.camel@citrix.com>
On Fri, 2015-09-18 at 11:27 +0100, Ian Campbell wrote:
> On Fri, 2015-09-18 at 10:02 +0100, Ian Campbell wrote:
> > On Wed, 2015-09-16 at 14:07 +0100, Ian Campbell wrote:
> > > [...]
> > > +sub add_event($$$$$) {
> > > + my ($job,$reso,$type,$info,$time) = @_;
> > > +
> > > + die unless $type =~ m/^(Start|End)/;
> > > +
> > > + $job->{OverallTime} //= {};
> > > +
> > > + $job->{Reso}{$reso} //= {};
> > > + $job->{Reso}{$reso}{$type} = $time;
> > > +
> > > + if ($type eq "Start") {
> > > + die if $job->{Reso}{$reso}{Info};
> >
> > This is attempting to check for conflicting events.
> >
> > However in the case of sharing hosts as noted in the commit message
> > there
> > is a gotcha which is that while the host is being prepared by one job
> > the
> > rest appears as rogue/anonymous jobs (with Info==(preparing), FWIW) and
> > I've just observed this happen because 2 of those had the same start
> > time
> > (which I suppose will be quite common?).
> >
> > The attached plan shows this for the sharing of 'host grain-weevil' for
> > "build-wheezy-i386 e1adb8b552dda5c653958083c9c7457f0398ad09" purposes.
> >
> > Anonymous jobs use $reso + $info as their (faked) job name, so they end
> > up
> > clashing here.
> >
> > I did the below as a quick fix, now I'm wondering if perhaps I should
> > suffix the faked up job name of these anon jobs with something unique
> > from
> > either $evt->{Share} or $evt->{Allocated}.
> >
> > ->{Allocated}{Shareix} looks promising. What do you think?
>
> This would be the following. So far I think this is the best option.
We discussed this IRL and it turns out for these sorts of "clashing" shares
the start and end should always be the same, and while ms-flights-summary
could check this it's not really its responsibility to do so (and that
failing to produce any output at all under these circumstances is totally
suboptimal).
Therefore I've dropped the sanity checks, as below. If ms-planner were to
go mad then there might be some anomalous looking rogue jobs in the table,
which is fine.
I'm also going to colourise the "Scheduled" column to indicate more clearly
when something is running right now.
Ian.
diff --git a/ms-flights-summary b/ms-flights-summary
index 964ca18..ba51eaa 100755
--- a/ms-flights-summary
+++ b/ms-flights-summary
@@ -152,13 +152,8 @@ sub add_event($$$$$) {
$job->{Reso}{$reso} //= {};
$job->{Reso}{$reso}{$type} = $time;
- if ($type eq "Start") {
- die "$job $reso $job->{Reso}{$reso}{Info}"
- if ($job->{Reso}{$reso}{Info}//$info) ne $info;
- $job->{Reso}{$reso}{Info} = $info;
- } else {
- die if $job->{Reso}{$reso}{Info} ne $info;
- }
+ $job->{Reso}{$reso}{Info} = $info
+ if $type eq "Start";
my $cmp = $type eq "Start" ?
sub { $_[0] < $_[1] ? $_[0] : $_[1] } :
prev parent reply other threads:[~2015-09-18 11:14 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-09-16 13:07 [PATCH OSSTEST v4] ms-flights-summary: Produce an HTML report of all active flights Ian Campbell
2015-09-17 16:14 ` Ian Campbell
2015-09-17 16:14 ` [PATCH OSSTEST 1/4] ms-queuedaemon: report-plan: Use rename-into-place for data-*.final.pl Ian Campbell
2015-09-17 16:47 ` Ian Campbell
2015-09-17 16:14 ` [PATCH OSSTEST 2/4] ms-queuedaemon: Break out catching-internally Ian Campbell
2015-09-17 16:47 ` Ian Campbell
2015-09-17 16:14 ` [PATCH OSSTEST 3/4] ms-queuedaemon: Add report-projection Ian Campbell
2015-09-17 16:33 ` Ian Jackson
2015-09-17 16:39 ` Ian Campbell
2015-09-17 16:14 ` [PATCH OSSTEST 4/4] ms-queuedaemon: Call ms-flights-summary upon completed plan/projection Ian Campbell
2015-09-17 16:34 ` Ian Jackson
2015-09-18 9:02 ` [PATCH OSSTEST v4] ms-flights-summary: Produce an HTML report of all active flights Ian Campbell
2015-09-18 10:27 ` Ian Campbell
2015-09-18 11:14 ` Ian Campbell [this message]
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=1442574865.18856.290.camel@citrix.com \
--to=ian.campbell@citrix.com \
--cc=ian.jackson@eu.citrix.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.