[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [Xen-devel] [PATCH] Minor fix to xentop to stop it dying when domains go away at the wrong time


  • To: "Keir Fraser" <Keir.Fraser@xxxxxxxxxxxx>
  • From: "Graham, Simon" <Simon.Graham@xxxxxxxxxxx>
  • Date: Tue, 25 Jul 2006 14:40:11 -0400
  • Cc: xen-devel@xxxxxxxxxxxxxxxxxxx
  • Delivery-date: Tue, 25 Jul 2006 11:41:33 -0700
  • List-id: Xen developer discussion <xen-devel.lists.xensource.com>
  • Thread-index: Acav/f/IGhDdGcXwRMuoGSnx6qmqzwAGxMVQ
  • Thread-topic: [Xen-devel] [PATCH] Minor fix to xentop to stop it dying when domains go away at the wrong time

> 
> libxenstat should be fixed so that xenstat_get_node() does not
> spuriously fail in this way. This could be done by getting the
> 'collectors' it calls to return better error info -- we can use EAGAIN
> to cause xenstat_get_node() to rerun itself from scratch rather than
> returning failure to the caller. Otherwise we're going to need to fix
> every user of libxenstat, which is stupid.
> 
>   -- Keir

I thought about this too but decided not to for a couple of reasons:

1. It's not clear that all possible users of xenstat_get_node() would
want
   to have the call take an arbitrary amount of time as it struggles to
   get a consistent snapshot -- better to let the caller decide policy
   on retrying the call.

2. xentop is currently the only user of xenstat_get_node in the tree and
the
   fix in xentop was waaay easier ;-)

If you still think xenstat_get_node() should loop until it has a
consistent snapshot
then I'll redo the patch that way

/simgr


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.