[CRIU] p.haul page statistics

[Mike Rapoport]

On Thu, Sep 08, 2016 at 09:34:11AM +0200, Adrian Reber wrote:
> On Thu, Sep 08, 2016 at 08:39:06AM +0300, Mike Rapoport wrote:
> > On Tue, Sep 06, 2016 at 11:35:23AM +0200, Adrian Reber wrote:
> > > Using the latest criu with p.haul the information about the pages dumped
> > > looks a bit wrong. From p.haul I get:
> > > 
> > > 09:27:59.485: 11602: 	Issuing pre-dump command to service
> > > 09:28:00.533: 11602: 	Dumped 13 pages, 0 skipped
> > > 09:28:00.533: 11602: Checking iteration progress:
> > > 09:28:00.533: 11602: 	> Small dump
> > > 
> > > But in the stats file I see:
> > > 
> > > {
> > >     "magic": "STATS", 
> > >     "entries": [
> > >         {
> > >             "dump": {
> > >                 "freezing_time": 101, 
> > >                 "frozen_time": 15680, 
> > >                 "memdump_time": 12033, 
> > >                 "memwrite_time": 1729763, 
> > >                 "pages_scanned": 154674, 
> > >                 "pages_skipped_parent": 0, 
> > >                 "pages_written": 13, 
> > >                 "irmap_resolve": 0, 
> > >                 "pages_zero": 0, 
> > >                 "pages_lazy": 153614
> > >             }
> > >         }
> > >     ]
> > > }
> > > 
> > > So p.haul is kind of right as the stats file says that only 13 pages
> > > have been written. But all the lazy pages have also been written.
> > > 
> > > Should pages_written include the lazy pages or should p.haul add
> > > pages_written and pages_lazy to get the actual number of pages written?
> > 
> > I think p.haul should add pages_written and pages_lazy.
> > 
> > When we will add post-copy support to p.haul, we will take care of proper
> > reporting for lazy pages.
> 
> Some quick fix for p.haul would be nice because right now the pre-dump
> iteration decision is based on pages_written which is probably 'wrong'
> most of the time when using the criu-dev branch.

I'm completely unfamiliar with p.haul, so if you could help with the fix,
it'll be great...
 
> 		Adrian
> 
--
Sincerely yours,
Mike.