| Message ID | 1564741121-1840-1-git-send-email-ivanren@tencent.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [v3] migration: always initial ram_counters for a new migration | expand |
On Fri, Aug 02, 2019 at 06:18:41PM +0800, Ivan Ren wrote: >From: Ivan Ren <ivanren@tencent.com> > >This patch fix a multifd migration bug in migration speed calculation, this >problem can be reproduced as follows: >1. start a vm and give a heavy memory write stress to prevent the vm be > successfully migrated to destination >2. begin a migration with multifd >3. migrate for a long time [actually, this can be measured by transferred bytes] >4. migrate cancel >5. begin a new migration with multifd, the migration will directly run into > migration_completion phase > >Reason as follows: > >Migration update bandwidth and s->threshold_size in function >migration_update_counters after BUFFER_DELAY time: > > current_bytes = migration_total_bytes(s); > transferred = current_bytes - s->iteration_initial_bytes; > time_spent = current_time - s->iteration_start_time; > bandwidth = (double)transferred / time_spent; > s->threshold_size = bandwidth * s->parameters.downtime_limit; > >In multifd migration, migration_total_bytes function return >qemu_ftell(s->to_dst_file) + ram_counters.multifd_bytes. >s->iteration_initial_bytes will be initialized to 0 at every new migration, >but ram_counters is a global variable, and history migration data will be >accumulated. So if the ram_counters.multifd_bytes is big enough, it may lead >pending_size >= s->threshold_size become false in migration_iteration_run >after the first migration_update_counters. > >Signed-off-by: Ivan Ren <ivanren@tencent.com> >Reviewed-by: Juan Quintela <quintela@redhat.com> >Suggested-by: Wei Yang <richardw.yang@linux.intel.com> >--- >v2->v3: >- fix the bug of update_iteration_initial_status function prototype > Code looks good. Have you verified on this version? BTW, you didn't address the multifd count in this patch, right?
On Mon, Aug 5, 2019 at 8:34 AM Wei Yang <richardw.yang@linux.intel.com> wrote: > > On Fri, Aug 02, 2019 at 06:18:41PM +0800, Ivan Ren wrote: > >From: Ivan Ren <ivanren@tencent.com> > > > >This patch fix a multifd migration bug in migration speed calculation, this > >problem can be reproduced as follows: > >1. start a vm and give a heavy memory write stress to prevent the vm be > > successfully migrated to destination > >2. begin a migration with multifd > >3. migrate for a long time [actually, this can be measured by transferred bytes] > >4. migrate cancel > >5. begin a new migration with multifd, the migration will directly run into > > migration_completion phase > > > >Reason as follows: > > > >Migration update bandwidth and s->threshold_size in function > >migration_update_counters after BUFFER_DELAY time: > > > > current_bytes = migration_total_bytes(s); > > transferred = current_bytes - s->iteration_initial_bytes; > > time_spent = current_time - s->iteration_start_time; > > bandwidth = (double)transferred / time_spent; > > s->threshold_size = bandwidth * s->parameters.downtime_limit; > > > >In multifd migration, migration_total_bytes function return > >qemu_ftell(s->to_dst_file) + ram_counters.multifd_bytes. > >s->iteration_initial_bytes will be initialized to 0 at every new migration, > >but ram_counters is a global variable, and history migration data will be > >accumulated. So if the ram_counters.multifd_bytes is big enough, it may lead > >pending_size >= s->threshold_size become false in migration_iteration_run > >after the first migration_update_counters. > > > >Signed-off-by: Ivan Ren <ivanren@tencent.com> > >Reviewed-by: Juan Quintela <quintela@redhat.com> > >Suggested-by: Wei Yang <richardw.yang@linux.intel.com> > >--- > >v2->v3: > >- fix the bug of update_iteration_initial_status function prototype > > > > Code looks good. Have you verified on this version? Yes > BTW, you didn't address the multifd count in this patch, right? Yes. Currently multifd page count has no harm, so I think it's better to optimize it in a new patch to make things clearer. Thanks.
On Mon, Aug 05, 2019 at 09:16:24AM +0800, Ivan Ren wrote: >On Mon, Aug 5, 2019 at 8:34 AM Wei Yang <richardw.yang@linux.intel.com> wrote: >> >> On Fri, Aug 02, 2019 at 06:18:41PM +0800, Ivan Ren wrote: >> >From: Ivan Ren <ivanren@tencent.com> >> > >> >This patch fix a multifd migration bug in migration speed calculation, this >> >problem can be reproduced as follows: >> >1. start a vm and give a heavy memory write stress to prevent the vm be >> > successfully migrated to destination >> >2. begin a migration with multifd >> >3. migrate for a long time [actually, this can be measured by transferred bytes] >> >4. migrate cancel >> >5. begin a new migration with multifd, the migration will directly run into >> > migration_completion phase >> > >> >Reason as follows: >> > >> >Migration update bandwidth and s->threshold_size in function >> >migration_update_counters after BUFFER_DELAY time: >> > >> > current_bytes = migration_total_bytes(s); >> > transferred = current_bytes - s->iteration_initial_bytes; >> > time_spent = current_time - s->iteration_start_time; >> > bandwidth = (double)transferred / time_spent; >> > s->threshold_size = bandwidth * s->parameters.downtime_limit; >> > >> >In multifd migration, migration_total_bytes function return >> >qemu_ftell(s->to_dst_file) + ram_counters.multifd_bytes. >> >s->iteration_initial_bytes will be initialized to 0 at every new migration, >> >but ram_counters is a global variable, and history migration data will be >> >accumulated. So if the ram_counters.multifd_bytes is big enough, it may lead >> >pending_size >= s->threshold_size become false in migration_iteration_run >> >after the first migration_update_counters. >> > >> >Signed-off-by: Ivan Ren <ivanren@tencent.com> >> >Reviewed-by: Juan Quintela <quintela@redhat.com> >> >Suggested-by: Wei Yang <richardw.yang@linux.intel.com> >> >--- >> >v2->v3: >> >- fix the bug of update_iteration_initial_status function prototype >> > >> >> Code looks good. Have you verified on this version? > >Yes > >> BTW, you didn't address the multifd count in this patch, right? > >Yes. >Currently multifd page count has no harm, so I think it's better to >optimize it in a new patch to make things clearer. Fine. Reviewed-by: Wei Yang <richardw.yang@linux.intel.com> > >Thanks.
* Ivan Ren (renyime@gmail.com) wrote: > From: Ivan Ren <ivanren@tencent.com> > > This patch fix a multifd migration bug in migration speed calculation, this > problem can be reproduced as follows: > 1. start a vm and give a heavy memory write stress to prevent the vm be > successfully migrated to destination > 2. begin a migration with multifd > 3. migrate for a long time [actually, this can be measured by transferred bytes] > 4. migrate cancel > 5. begin a new migration with multifd, the migration will directly run into > migration_completion phase > > Reason as follows: > > Migration update bandwidth and s->threshold_size in function > migration_update_counters after BUFFER_DELAY time: > > current_bytes = migration_total_bytes(s); > transferred = current_bytes - s->iteration_initial_bytes; > time_spent = current_time - s->iteration_start_time; > bandwidth = (double)transferred / time_spent; > s->threshold_size = bandwidth * s->parameters.downtime_limit; > > In multifd migration, migration_total_bytes function return > qemu_ftell(s->to_dst_file) + ram_counters.multifd_bytes. > s->iteration_initial_bytes will be initialized to 0 at every new migration, > but ram_counters is a global variable, and history migration data will be > accumulated. So if the ram_counters.multifd_bytes is big enough, it may lead > pending_size >= s->threshold_size become false in migration_iteration_run > after the first migration_update_counters. > > Signed-off-by: Ivan Ren <ivanren@tencent.com> > Reviewed-by: Juan Quintela <quintela@redhat.com> > Suggested-by: Wei Yang <richardw.yang@linux.intel.com> Thank you, Queued > --- > v2->v3: > - fix the bug of update_iteration_initial_status function prototype > > v1->v2: > - Add interface update_iteration_initial_status to update statistic fields > at the same time to avoid info mismatch lead wrong calculation result. > > migration/migration.c | 25 +++++++++++++++++++------ > migration/savevm.c | 1 + > 2 files changed, 20 insertions(+), 6 deletions(-) > > diff --git a/migration/migration.c b/migration/migration.c > index 8a607fe1e2..bea9b1d796 100644 > --- a/migration/migration.c > +++ b/migration/migration.c > @@ -1908,6 +1908,11 @@ static bool migrate_prepare(MigrationState *s, bool blk, bool blk_inc, > } > > migrate_init(s); > + /* > + * set ram_counters memory to zero for a > + * new migration > + */ > + memset(&ram_counters, 0, sizeof(ram_counters)); > > return true; > } > @@ -3025,6 +3030,17 @@ static void migration_calculate_complete(MigrationState *s) > } > } > > +static void update_iteration_initial_status(MigrationState *s) > +{ > + /* > + * Update these three fields at the same time to avoid mismatch info lead > + * wrong speed calculation. > + */ > + s->iteration_start_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME); > + s->iteration_initial_bytes = migration_total_bytes(s); > + s->iteration_initial_pages = ram_get_total_transferred_pages(); > +} > + > static void migration_update_counters(MigrationState *s, > int64_t current_time) > { > @@ -3060,9 +3076,7 @@ static void migration_update_counters(MigrationState *s, > > qemu_file_reset_rate_limit(s->to_dst_file); > > - s->iteration_start_time = current_time; > - s->iteration_initial_bytes = current_bytes; > - s->iteration_initial_pages = ram_get_total_transferred_pages(); > + update_iteration_initial_status(s); > > trace_migrate_transferred(transferred, time_spent, > bandwidth, s->threshold_size); > @@ -3186,7 +3200,7 @@ static void *migration_thread(void *opaque) > rcu_register_thread(); > > object_ref(OBJECT(s)); > - s->iteration_start_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME); > + update_iteration_initial_status(s); > > qemu_savevm_state_header(s->to_dst_file); > > @@ -3251,8 +3265,7 @@ static void *migration_thread(void *opaque) > * the local variables. This is important to avoid > * breaking transferred_bytes and bandwidth calculation > */ > - s->iteration_start_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME); > - s->iteration_initial_bytes = 0; > + update_iteration_initial_status(s); > } > > current_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME); > diff --git a/migration/savevm.c b/migration/savevm.c > index 79ed44d475..480c511b19 100644 > --- a/migration/savevm.c > +++ b/migration/savevm.c > @@ -1424,6 +1424,7 @@ static int qemu_savevm_state(QEMUFile *f, Error **errp) > } > > migrate_init(ms); > + memset(&ram_counters, 0, sizeof(ram_counters)); > ms->to_dst_file = f; > > qemu_mutex_unlock_iothread(); > -- > 2.17.2 (Apple Git-113) > > -- Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK
diff --git a/migration/migration.c b/migration/migration.c index 8a607fe1e2..bea9b1d796 100644 --- a/migration/migration.c +++ b/migration/migration.c @@ -1908,6 +1908,11 @@ static bool migrate_prepare(MigrationState *s, bool blk, bool blk_inc, } migrate_init(s); + /* + * set ram_counters memory to zero for a + * new migration + */ + memset(&ram_counters, 0, sizeof(ram_counters)); return true; } @@ -3025,6 +3030,17 @@ static void migration_calculate_complete(MigrationState *s) } } +static void update_iteration_initial_status(MigrationState *s) +{ + /* + * Update these three fields at the same time to avoid mismatch info lead + * wrong speed calculation. + */ + s->iteration_start_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME); + s->iteration_initial_bytes = migration_total_bytes(s); + s->iteration_initial_pages = ram_get_total_transferred_pages(); +} + static void migration_update_counters(MigrationState *s, int64_t current_time) { @@ -3060,9 +3076,7 @@ static void migration_update_counters(MigrationState *s, qemu_file_reset_rate_limit(s->to_dst_file); - s->iteration_start_time = current_time; - s->iteration_initial_bytes = current_bytes; - s->iteration_initial_pages = ram_get_total_transferred_pages(); + update_iteration_initial_status(s); trace_migrate_transferred(transferred, time_spent, bandwidth, s->threshold_size); @@ -3186,7 +3200,7 @@ static void *migration_thread(void *opaque) rcu_register_thread(); object_ref(OBJECT(s)); - s->iteration_start_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME); + update_iteration_initial_status(s); qemu_savevm_state_header(s->to_dst_file); @@ -3251,8 +3265,7 @@ static void *migration_thread(void *opaque) * the local variables. This is important to avoid * breaking transferred_bytes and bandwidth calculation */ - s->iteration_start_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME); - s->iteration_initial_bytes = 0; + update_iteration_initial_status(s); } current_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME); diff --git a/migration/savevm.c b/migration/savevm.c index 79ed44d475..480c511b19 100644 --- a/migration/savevm.c +++ b/migration/savevm.c @@ -1424,6 +1424,7 @@ static int qemu_savevm_state(QEMUFile *f, Error **errp) } migrate_init(ms); + memset(&ram_counters, 0, sizeof(ram_counters)); ms->to_dst_file = f; qemu_mutex_unlock_iothread();