[v8,3/8] tests: Add migration xbzrle test

Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
---
 tests/migration-test.c | 64 ++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 64 insertions(+)

* Juan Quintela (quintela@redhat.com) wrote:
> Signed-off-by: Juan Quintela <quintela@redhat.com>
> Reviewed-by: Peter Xu <peterx@redhat.com>
> ---
>  tests/migration-test.c | 64 ++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 64 insertions(+)
> 
> diff --git a/tests/migration-test.c b/tests/migration-test.c
> index 834cdf50f2..fd885ba909 100644
> --- a/tests/migration-test.c
> +++ b/tests/migration-test.c
> @@ -512,6 +512,20 @@ static void deprecated_set_speed(QTestState *who, const char *value)
>      migrate_check_parameter(who, "max-bandwidth", value);
>  }
>  
> +static void deprecated_set_cache_size(QTestState *who, const char *value)
> +{
> +    QDict *rsp;
> +    gchar *cmd;
> +
> +    cmd = g_strdup_printf("{ 'execute': 'migrate-set-cache-size',"
> +                          "'arguments': { 'value': %s } }", value);
> +    rsp = qtest_qmp(who, cmd);
> +    g_free(cmd);
> +    g_assert(qdict_haskey(rsp, "return"));
> +    QDECREF(rsp);
> +    migrate_check_parameter(who, "xbzrle-cache-size", value);
> +}
> +
>  static void test_deprecated(void)
>  {
>      QTestState *from;
> @@ -520,6 +534,7 @@ static void test_deprecated(void)
>  
>      deprecated_set_downtime(from, 0.12345);
>      deprecated_set_speed(from, "12345");
> +    deprecated_set_cache_size(from, "4096");
>  
>      qtest_quit(from);
>  }
> @@ -634,6 +649,54 @@ static void test_precopy_unix(void)
>      g_free(uri);
>  }
>  
> +static void test_xbzrle(const char *uri)
> +{
> +    QTestState *from, *to;
> +
> +    test_migrate_start(&from, &to, uri, false);
> +
> +    /* We want to pick a speed slow enough that the test completes
> +     * quickly, but that it doesn't complete precopy even on a slow
> +     * machine, so also set the downtime.
> +     */
> +    /* 1 ms should make it not converge*/
> +    migrate_set_parameter(from, "downtime-limit", "1");
> +    /* 1GB/s */
> +    migrate_set_parameter(from, "max-bandwidth", "1000000000");
> +
> +    migrate_set_parameter(from, "xbzrle-cache-size", "33554432");

I still worry about the cache size relative to the size of memory we're
actually changing in the test; I don't quite understand why it's turning
out to get lots of hits.

Also, xbzrle eats so much CPU we'll still have to watch out for the low
end CPUs.

Dave

> +    migrate_set_capability(from, "xbzrle", "true");
> +    migrate_set_capability(to, "xbzrle", "true");
> +    /* Wait for the first serial output from the source */
> +    wait_for_serial("src_serial");
> +
> +    migrate(from, uri);
> +
> +    wait_for_migration_pass(from);
> +
> +    /* 300ms should converge */
> +    migrate_set_parameter(from, "downtime-limit", "300");
> +
> +    if (!got_stop) {
> +        qtest_qmp_eventwait(from, "STOP");
> +    }
> +    qtest_qmp_eventwait(to, "RESUME");
> +
> +    wait_for_serial("dest_serial");
> +    wait_for_migration_complete(from);
> +
> +    test_migrate_end(from, to, true);
> +}
> +
> +static void test_xbzrle_unix(void)
> +{
> +    char *uri = g_strdup_printf("unix:%s/migsocket", tmpfs);
> +
> +    test_xbzrle(uri);
> +    g_free(uri);
> +}
> +
>  int main(int argc, char **argv)
>  {
>      char template[] = "/tmp/migration-test-XXXXXX";
> @@ -657,6 +720,7 @@ int main(int argc, char **argv)
>      qtest_add_func("/migration/deprecated", test_deprecated);
>      qtest_add_func("/migration/bad_dest", test_baddest);
>      qtest_add_func("/migration/precopy/unix", test_precopy_unix);
> +    qtest_add_func("/migration/xbzrle/unix", test_xbzrle_unix);
>  
>      ret = g_test_run();
>  
> -- 
> 2.17.0
> 
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

"Dr. David Alan Gilbert" <dgilbert@redhat.com> wrote:
> * Juan Quintela (quintela@redhat.com) wrote:
>> Signed-off-by: Juan Quintela <quintela@redhat.com>
>> Reviewed-by: Peter Xu <peterx@redhat.com>

>> +
>> +    migrate_set_parameter(from, "xbzrle-cache-size", "33554432");
>
> I still worry about the cache size relative to the size of memory we're
> actually changing in the test; I don't quite understand why it's turning
> out to get lots of hits.

I planned this as a smorke test.  But what value do you have in mind?

> Also, xbzrle eats so much CPU we'll still have to watch out for the low
> end CPUs.

Any concrete suggestion?

My plan would have been to push on my next pull request:

Add migration precopy test
Add migration xbzrle test
Migration ppc now inlines its program

And put the rest of the patches with fixes for another review.


Later, Juan.

* Juan Quintela (quintela@redhat.com) wrote:
> "Dr. David Alan Gilbert" <dgilbert@redhat.com> wrote:
> > * Juan Quintela (quintela@redhat.com) wrote:
> >> Signed-off-by: Juan Quintela <quintela@redhat.com>
> >> Reviewed-by: Peter Xu <peterx@redhat.com>
> 
> >> +
> >> +    migrate_set_parameter(from, "xbzrle-cache-size", "33554432");
> >
> > I still worry about the cache size relative to the size of memory we're
> > actually changing in the test; I don't quite understand why it's turning
> > out to get lots of hits.
> 
> I planned this as a smorke test.  But what value do you have in mind?

Given the test code dirties 100MB of RAM, I suggest a cache of 110MB
say.  But it's interesting that the stats show we are getting a lot of
page hits; I'm not sure I understand why.

> > Also, xbzrle eats so much CPU we'll still have to watch out for the low
> > end CPUs.
> 
> Any concrete suggestion?
> 
> My plan would have been to push on my next pull request:
> 
> Add migration precopy test
> Add migration xbzrle test
> Migration ppc now inlines its program
> 
> And put the rest of the patches with fixes for another review.

Go with it and lets see; if we hit problems then we can just turn that
300ms downtime limit right up.

Dave

> 
> Later, Juan.
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK