[Qemu-devel] [PATCH 5/7] migration: Fix the migrate auto converge process

[Qemu-devel] [PATCH 5/7] migration: Fix the migrate auto converge process

[Gonglei (Arei)]

It is inaccuracy and complex that using the transfer speed of
migration thread to determine whether the convergence migration.
The dirty page may be compressed by XBZRLE or ZERO_PAGE.The counter
of updating dirty bitmap will be increasing continuously if the
migration can't convergence.

Signed-off-by: ChenLiang <chenliang88@huawei.com>
Signed-off-by: Gonglei <arei.gonglei@huawei.com>
---
 arch_init.c | 26 +++-----------------------
 1 file changed, 3 insertions(+), 23 deletions(-)

diff --git a/arch_init.c b/arch_init.c
index fc71331..2211e0b 100644
--- a/arch_init.c
+++ b/arch_init.c
@@ -107,7 +107,6 @@ int graphic_depth = 32;
 
 const uint32_t arch_type = QEMU_ARCH;
 static bool mig_throttle_on;
-static int dirty_rate_high_cnt;
 static void check_guest_throttling(void);
 
 static uint64_t bitmap_sync_cnt;
@@ -464,17 +463,11 @@ static void migration_bitmap_sync(void)
     uint64_t num_dirty_pages_init = migration_dirty_pages;
     MigrationState *s = migrate_get_current();
     static int64_t start_time;
-    static int64_t bytes_xfer_prev;
     static int64_t num_dirty_pages_period;
     int64_t end_time;
-    int64_t bytes_xfer_now;
 
     increase_bitmap_sync_cnt();
 
-    if (!bytes_xfer_prev) {
-        bytes_xfer_prev = ram_bytes_transferred();
-    }
-
     if (!start_time) {
         start_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME);
     }
@@ -493,21 +486,9 @@ static void migration_bitmap_sync(void)
     /* more than 1 second = 1000 millisecons */
     if (end_time > start_time + 1000) {
         if (migrate_auto_converge()) {
-            /* The following detection logic can be refined later. For now:
-               Check to see if the dirtied bytes is 50% more than the approx.
-               amount of bytes that just got transferred since the last time we
-               were in this routine. If that happens >N times (for now N==4)
-               we turn on the throttle down logic */
-            bytes_xfer_now = ram_bytes_transferred();
-            if (s->dirty_pages_rate &&
-               (num_dirty_pages_period * TARGET_PAGE_SIZE >
-                   (bytes_xfer_now - bytes_xfer_prev)/2) &&
-               (dirty_rate_high_cnt++ > 4)) {
-                    trace_migration_throttle();
-                    mig_throttle_on = true;
-                    dirty_rate_high_cnt = 0;
-             }
-             bytes_xfer_prev = bytes_xfer_now;
+            if (get_bitmap_sync_cnt() > 15) {
+                mig_throttle_on = true;
+            }
         } else {
              mig_throttle_on = false;
         }
@@ -712,7 +693,6 @@ static int ram_save_setup(QEMUFile *f, void *opaque)
     bitmap_set(migration_bitmap, 0, ram_pages);
     migration_dirty_pages = ram_pages;
     mig_throttle_on = false;
-    dirty_rate_high_cnt = 0;
     reset_bitmap_sync_cnt();
 
     if (migrate_use_xbzrle()) {
-- 
1.7.12.4

Best regards,
-Gonglei

Re: [Qemu-devel] [PATCH 5/7] migration: Fix the migrate auto converge process

[Dr. David Alan Gilbert]

* Gonglei (Arei) (arei.gonglei@huawei.com) wrote:
> It is inaccuracy and complex that using the transfer speed of
> migration thread to determine whether the convergence migration.
> The dirty page may be compressed by XBZRLE or ZERO_PAGE.The counter
> of updating dirty bitmap will be increasing continuously if the
> migration can't convergence.
> 
> Signed-off-by: ChenLiang <chenliang88@huawei.com>
> Signed-off-by: Gonglei <arei.gonglei@huawei.com>
> ---
>  arch_init.c | 26 +++-----------------------
>  1 file changed, 3 insertions(+), 23 deletions(-)
> 
> diff --git a/arch_init.c b/arch_init.c
> index fc71331..2211e0b 100644
> --- a/arch_init.c
> +++ b/arch_init.c
> @@ -107,7 +107,6 @@ int graphic_depth = 32;
>  
>  const uint32_t arch_type = QEMU_ARCH;
>  static bool mig_throttle_on;
> -static int dirty_rate_high_cnt;
>  static void check_guest_throttling(void);
>  
>  static uint64_t bitmap_sync_cnt;
> @@ -464,17 +463,11 @@ static void migration_bitmap_sync(void)
>      uint64_t num_dirty_pages_init = migration_dirty_pages;
>      MigrationState *s = migrate_get_current();
>      static int64_t start_time;
> -    static int64_t bytes_xfer_prev;
>      static int64_t num_dirty_pages_period;
>      int64_t end_time;
> -    int64_t bytes_xfer_now;
>  
>      increase_bitmap_sync_cnt();
>  
> -    if (!bytes_xfer_prev) {
> -        bytes_xfer_prev = ram_bytes_transferred();
> -    }
> -
>      if (!start_time) {
>          start_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME);
>      }
> @@ -493,21 +486,9 @@ static void migration_bitmap_sync(void)
>      /* more than 1 second = 1000 millisecons */
>      if (end_time > start_time + 1000) {
>          if (migrate_auto_converge()) {
> -            /* The following detection logic can be refined later. For now:
> -               Check to see if the dirtied bytes is 50% more than the approx.
> -               amount of bytes that just got transferred since the last time we
> -               were in this routine. If that happens >N times (for now N==4)
> -               we turn on the throttle down logic */
> -            bytes_xfer_now = ram_bytes_transferred();
> -            if (s->dirty_pages_rate &&
> -               (num_dirty_pages_period * TARGET_PAGE_SIZE >
> -                   (bytes_xfer_now - bytes_xfer_prev)/2) &&
> -               (dirty_rate_high_cnt++ > 4)) {
> -                    trace_migration_throttle();
> -                    mig_throttle_on = true;
> -                    dirty_rate_high_cnt = 0;
> -             }
> -             bytes_xfer_prev = bytes_xfer_now;
> +            if (get_bitmap_sync_cnt() > 15) {
> +                mig_throttle_on = true;
> +            }

That is a lot simpler, and I suspect as good - again I'd
move that magic '15' to a constant somewhere.

What have you tested this on - have you tested with really big RAM VMs?
What's it's behaviour like with rate-limiting?

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

Re: [Qemu-devel] [PATCH 5/7] migration: Fix the migrate auto converge process

[Gonglei]

On 2014/2/28 18:16, Dr. David Alan Gilbert wrote:

> * Gonglei (Arei) (arei.gonglei@huawei.com) wrote:
>> It is inaccuracy and complex that using the transfer speed of
>> migration thread to determine whether the convergence migration.
>> The dirty page may be compressed by XBZRLE or ZERO_PAGE.The counter
>> of updating dirty bitmap will be increasing continuously if the
>> migration can't convergence.
>>
>> Signed-off-by: ChenLiang <chenliang88@huawei.com>
>> Signed-off-by: Gonglei <arei.gonglei@huawei.com>
>> ---
>>  arch_init.c | 26 +++-----------------------
>>  1 file changed, 3 insertions(+), 23 deletions(-)
>>
>> diff --git a/arch_init.c b/arch_init.c
>> index fc71331..2211e0b 100644
>> --- a/arch_init.c
>> +++ b/arch_init.c
>> @@ -107,7 +107,6 @@ int graphic_depth = 32;
>>  
>>  const uint32_t arch_type = QEMU_ARCH;
>>  static bool mig_throttle_on;
>> -static int dirty_rate_high_cnt;
>>  static void check_guest_throttling(void);
>>  
>>  static uint64_t bitmap_sync_cnt;
>> @@ -464,17 +463,11 @@ static void migration_bitmap_sync(void)
>>      uint64_t num_dirty_pages_init = migration_dirty_pages;
>>      MigrationState *s = migrate_get_current();
>>      static int64_t start_time;
>> -    static int64_t bytes_xfer_prev;
>>      static int64_t num_dirty_pages_period;
>>      int64_t end_time;
>> -    int64_t bytes_xfer_now;
>>  
>>      increase_bitmap_sync_cnt();
>>  
>> -    if (!bytes_xfer_prev) {
>> -        bytes_xfer_prev = ram_bytes_transferred();
>> -    }
>> -
>>      if (!start_time) {
>>          start_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME);
>>      }
>> @@ -493,21 +486,9 @@ static void migration_bitmap_sync(void)
>>      /* more than 1 second = 1000 millisecons */
>>      if (end_time > start_time + 1000) {
>>          if (migrate_auto_converge()) {
>> -            /* The following detection logic can be refined later. For now:
>> -               Check to see if the dirtied bytes is 50% more than the approx.
>> -               amount of bytes that just got transferred since the last time we
>> -               were in this routine. If that happens >N times (for now N==4)
>> -               we turn on the throttle down logic */
>> -            bytes_xfer_now = ram_bytes_transferred();
>> -            if (s->dirty_pages_rate &&
>> -               (num_dirty_pages_period * TARGET_PAGE_SIZE >
>> -                   (bytes_xfer_now - bytes_xfer_prev)/2) &&
>> -               (dirty_rate_high_cnt++ > 4)) {
>> -                    trace_migration_throttle();
>> -                    mig_throttle_on = true;
>> -                    dirty_rate_high_cnt = 0;
>> -             }
>> -             bytes_xfer_prev = bytes_xfer_now;
>> +            if (get_bitmap_sync_cnt() > 15) {
>> +                mig_throttle_on = true;
>> +            }
> 
> That is a lot simpler, and I suspect as good - again I'd
> move that magic '15' to a constant somewhere.
> 

Thanks, Check it.

> What have you tested this on - have you tested with really big RAM VMs?
> What's it's behaviour like with rate-limiting?
> 
> Dave

Yeah, We have tested it using 25G memory VM as the first mail said.

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


Best regards,
-Gonglei