[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-devel] [PATCH v2 3/8] migration: show the statistics of compre
From: |
Dr. David Alan Gilbert |
Subject: |
Re: [Qemu-devel] [PATCH v2 3/8] migration: show the statistics of compression |
Date: |
Wed, 25 Jul 2018 17:44:02 +0100 |
User-agent: |
Mutt/1.10.0 (2018-05-17) |
* Peter Xu (address@hidden) wrote:
> On Mon, Jul 23, 2018 at 03:39:18PM +0800, Xiao Guangrong wrote:
> >
> >
> > On 07/23/2018 12:36 PM, Peter Xu wrote:
> > > On Thu, Jul 19, 2018 at 08:15:15PM +0800, address@hidden wrote:
> > > > @@ -1597,6 +1608,24 @@ static void migration_update_rates(RAMState *rs,
> > > > int64_t end_time)
> > > > rs->xbzrle_cache_miss_prev) / iter_count;
> > > > rs->xbzrle_cache_miss_prev = xbzrle_counters.cache_miss;
> > > > }
> > > > +
> > > > + if (migrate_use_compression()) {
> > > > + uint64_t comp_pages;
> > > > +
> > > > + compression_counters.busy_rate =
> > > > (double)(compression_counters.busy -
> > > > + rs->compress_thread_busy_prev) / iter_count;
> > >
> > > Here I'm not sure it's correct...
> > >
> > > "iter_count" stands for ramstate.iterations. It's increased per
> > > ram_find_and_save_block(), so IMHO it might contain multiple guest
> >
> > ram_find_and_save_block() returns if a page is successfully posted and
> > it only posts 1 page out at one time.
>
> ram_find_and_save_block() calls ram_save_host_page(), and we should be
> sending multiple guest pages in ram_save_host_page() if the host page
> is a huge page?
>
> >
> > > pages. However compression_counters.busy should be per guest page.
> > >
> >
> > Actually, it's derived from xbzrle_counters.cache_miss_rate:
> > xbzrle_counters.cache_miss_rate =
> > (double)(xbzrle_counters.cache_miss -
> > rs->xbzrle_cache_miss_prev) / iter_count;
>
> Then this is suspecious to me too...
Actually; I think this isn't totally wrong; iter_count is the *difference* in
iterations since the last time it was updated:
uint64_t iter_count = rs->iterations - rs->iterations_prev;
xbzrle_counters.cache_miss_rate = (double)(xbzrle_counters.cache_miss -
rs->xbzrle_cache_miss_prev) / iter_count;
so this is:
cache-misses-since-last-update
------------------------------
iterations since last-update
so the 'miss_rate' is ~misses / iteration.
Although that doesn't really correspond to time.
Dave
> >
> > > > + rs->compress_thread_busy_prev = compression_counters.busy;
> > > > +
> > > > + comp_pages = compression_counters.pages -
> > > > rs->compress_pages_prev;
> > > > + if (comp_pages) {
> > > > + compression_counters.compression_rate =
> > > > + (double)(compression_counters.reduced_size -
> > > > + rs->compress_reduced_size_prev) /
> > > > + (comp_pages * TARGET_PAGE_SIZE);
> > > > + rs->compress_pages_prev = compression_counters.pages;
> > > > + rs->compress_reduced_size_prev =
> > > > compression_counters.reduced_size;
> > > > + }
> > > > + }
> > > > }
> > > > static void migration_bitmap_sync(RAMState *rs)
> > > > @@ -1872,6 +1901,9 @@ static void flush_compressed_data(RAMState *rs)
> > > > qemu_mutex_lock(&comp_param[idx].mutex);
> > > > if (!comp_param[idx].quit) {
> > > > len = qemu_put_qemu_file(rs->f, comp_param[idx].file);
> > > > + /* 8 means a header with RAM_SAVE_FLAG_CONTINUE. */
> > > > + compression_counters.reduced_size += TARGET_PAGE_SIZE -
> > > > len + 8;
> > >
> > > I would agree with Dave here - why we store the "reduced size" instead
> > > of the size of the compressed data (which I think should be len - 8)?
> > >
> >
> > len-8 is the size of data after compressed rather than the data improved
> > by compression that is not straightforward for the user to see how much
> > the improvement is by applying compression.
> >
> > Hmm... but it is not a big deal to me... :)
>
> Yeah it might be a personal preference indeed. :)
>
> It's just natural to do that this way for me since AFAIU the
> compression ratio is defined as:
>
> compressed data size
> compression ratio = ------------------------
> original data size
>
> >
> > > Meanwhile, would a helper be nicer? Like:
> >
> > Yup, that's nicer indeed.
>
> Regards,
>
> --
> Peter Xu
--
Dr. David Alan Gilbert / address@hidden / Manchester, UK
- Re: [Qemu-devel] [PATCH v2 1/8] migration: do not wait for free thread, (continued)
- [Qemu-devel] [PATCH v2 2/8] migration: fix counting normal page for compression, guangrong . xiao, 2018/07/19
- [Qemu-devel] [PATCH v2 3/8] migration: show the statistics of compression, guangrong . xiao, 2018/07/19
- Re: [Qemu-devel] [PATCH v2 3/8] migration: show the statistics of compression, Peter Xu, 2018/07/23
- Re: [Qemu-devel] [PATCH v2 3/8] migration: show the statistics of compression, Xiao Guangrong, 2018/07/23
- Re: [Qemu-devel] [PATCH v2 3/8] migration: show the statistics of compression, Peter Xu, 2018/07/23
- Re: [Qemu-devel] [PATCH v2 3/8] migration: show the statistics of compression, Xiao Guangrong, 2018/07/23
- Re: [Qemu-devel] [PATCH v2 3/8] migration: show the statistics of compression, Peter Xu, 2018/07/23
- Re: [Qemu-devel] [PATCH v2 3/8] migration: show the statistics of compression, Xiao Guangrong, 2018/07/24
- Re: [Qemu-devel] [PATCH v2 3/8] migration: show the statistics of compression,
Dr. David Alan Gilbert <=
- Re: [Qemu-devel] [PATCH v2 3/8] migration: show the statistics of compression, Peter Xu, 2018/07/26
[Qemu-devel] [PATCH v2 4/8] migration: introduce save_zero_page_to_file, guangrong . xiao, 2018/07/19
[Qemu-devel] [PATCH v2 5/8] migration: drop the return value of do_compress_ram_page, guangrong . xiao, 2018/07/19
[Qemu-devel] [PATCH v2 6/8] migration: move handle of zero page to the thread, guangrong . xiao, 2018/07/19