Hi Suiheng,
Sorry for slow reply, the document 'COLO-FT.txt' in qemu is out of date, I
will update it lately.
Please follow this step to run COLO(the command has been changed).
https://wiki.qemu.org/Features/COLO
Thanks
Zhang Chen
On Fri, Mar 9, 2018 at 10:54 AM,
From: zhanghailiang
COLO thread may sleep at qemu_sem_wait(&s->colo_checkpoint_sem),
while failover works begin, It's better to wakeup it to quick
the process.
Signed-off-by: zhanghailiang
---
migration/colo.c | 8
1 file changed, 8 insertions(+)
diff --git a/migration/colo.c b/migra
On Sun, Mar 11, 2018 at 5:33 AM, Eric Blake wrote:
> On 03/10/2018 03:22 PM, Zhang Chen wrote:
>
>> Libvirt or other high level sofware can use this command query colo
>> status.
>>
>
> s/sofware/software/
>
> You can test this command like that:
>> {&
From: zhanghailiang
Notify all net filters about the checkpoint and failover event.
Signed-off-by: zhanghailiang
---
migration/colo.c | 12
1 file changed, 12 insertions(+)
diff --git a/migration/colo.c b/migration/colo.c
index f3d588e..1a59baa 100644
--- a/migration/colo.c
+++ b
Filter needs to process the event of checkpoint/failover or
other event passed by COLO frame.
Signed-off-by: zhanghailiang
---
include/net/filter.h | 5 +
net/filter.c | 17 +
net/net.c| 28
3 files changed, 50 insertions(+)
: zhanghailiang
Signed-off-by: Zhang Chen
---
migration/colo.c | 13 +
net/filter-rewriter.c | 40
2 files changed, 53 insertions(+)
diff --git a/migration/colo.c b/migration/colo.c
index f514492..f3d588e 100644
--- a/migration/colo.c
+++ b
_main() and qemu_save_device_state()
public, and simplify the codes of qemu_save_device_state() by calling the
wrapper qemu_savevm_state_header().
Signed-off-by: zhanghailiang
Signed-off-by: Li Zhijian
Signed-off-by: Zhang Chen
Reviewed-by: Dr. David Alan Gilbert
---
migration/colo.c
From: zhanghailiang
Don't need to flush all VM's ram from cache, only
flush the dirty pages since last checkpoint
Signed-off-by: Li Zhijian
Signed-off-by: Zhang Chen
Signed-off-by: zhanghailiang
---
migration/ram.c | 12
1 file changed, 12 insertions(+)
diff --git a
Libvirt or other high level software can use this command query colo status.
You can test this command like that:
{'execute':'query-colo-status'}
Signed-off-by: Zhang Chen
---
migration/colo.c| 31 +++
qapi/migration.json | 33
-by: zhanghailiang
Signed-off-by: Zhang Chen
Reviewed-by: Dr. David Alan Gilbert
---
migration/colo.c | 42 --
migration/migration.c | 4 ++--
2 files changed, 42 insertions(+), 4 deletions(-)
diff --git a/migration/colo.c b/migration/colo.c
index
x27;s failover verdict,
it is still necessary to notify users that we exited COLO mode.
Signed-off-by: zhanghailiang
Signed-off-by: Li Zhijian
Signed-off-by: Zhang Chen
Reviewed-by: Eric Blake
---
migration/colo.c| 20
qapi/migration.json | 35 +
From: zhanghailiang
During the time of VM's running, PVM may dirty some pages, we will transfer
PVM's dirty pages to SVM and store them into SVM's RAM cache at next checkpoint
time. So, the content of SVM's RAM cache will always be same with PVM's memory
after checkpoint.
Instead of flushing all
Make sure master start block replication after slave's block
replication started.
Besides, we need to activate VM's blocks before goes into
COLO state.
Signed-off-by: zhanghailiang
Signed-off-by: Li Zhijian
Signed-off-by: Zhang Chen
---
migration/colo.c
eive all PVM's state.
Signed-off-by: zhanghailiang
Signed-off-by: Li Zhijian
Signed-off-by: Zhang Chen
---
include/exec/ram_addr.h | 1 +
migration/migration.c | 2 +
migration/ram.c | 99 +++--
migration/ram.h | 4 ++
mig
While do checkpoint, we need to flush all the unhandled packets,
By using the filter notifier mechanism, we can easily to notify
every compare object to do this process, which runs inside
of compare threads as a coroutine.
Signed-off-by: zhanghailiang
Signed-off-by: Zhang Chen
---
include
We record the address of the dirty pages that received,
it will help flushing pages that cached into SVM.
Here, it is a trick, we record dirty pages by re-using migration
dirty bitmap. In the later patch, we will start the dirty log
for SVM, just like migration, in this way, we can record both
the
After a net connection is closed, we didn't clear its releated resources
in connection_track_table, which will lead to memory leak.
Let't track the state of net connection, if it is closed, its related
resources will be cleared up.
Signed-off-by: zhanghailiang
Signed-off-by:
: zhanghailiang
Signed-off-by: Zhang Chen
---
include/migration/colo.h | 5 ++--
migration/Makefile.objs | 2 +-
migration/colo-comm.c| 76
migration/colo.c | 13 -
migration/migration.c| 23 ++-
migration/savevm.c
k part to indepentent patch set.
- Optimize codes.
Zhang Chen (10):
filter-rewriter: fix memory leak for connection in
connection_track_table
colo-compare: implement the process of checkpoint
colo-compare: use notifier to notify packets comparing result
COLO: integrate colo compare with
It's a good idea to use notifier to notify COLO frame of
inconsistent packets comparing.
Signed-off-by: Zhang Chen
Signed-off-by: zhanghailiang
---
net/colo-compare.c | 32 +---
net/colo-compare.h | 2 ++
2 files changed, 27 insertions(+), 7 deletions(-)
am codes.
- Fix mingw compile error in patch 02/16.
- Fix some comments.
- Fix conflect with the patch "migration: remove "enable_colo" var" in patch
06/16.
V3:
- Address community comments from V2.
- Rebase on upstream codes.
- Fix several bugs.
- Splite shared disk part to
From: zhanghailiang
Don't need to flush all VM's ram from cache, only
flush the dirty pages since last checkpoint
Signed-off-by: Li Zhijian
Signed-off-by: Zhang Chen
Signed-off-by: zhanghailiang
---
migration/ram.c | 12
1 file changed, 12 insertions(+)
diff --git a
From: zhanghailiang
Notify all net filters about the checkpoint and failover event.
Signed-off-by: zhanghailiang
---
migration/colo.c | 12
1 file changed, 12 insertions(+)
diff --git a/migration/colo.c b/migration/colo.c
index 6834f3d..5fb4d8c 100644
--- a/migration/colo.c
+++ b
From: zhanghailiang
COLO thread may sleep at qemu_sem_wait(&s->colo_checkpoint_sem),
while failover works begin, It's better to wakeup it to quick
the process.
Signed-off-by: zhanghailiang
---
migration/colo.c | 8
1 file changed, 8 insertions(+)
diff --git a/migration/colo.c b/migra
Libvirt or other high level sofware can use this command query colo status.
You can test this command like that:
{'execute':'query-colo-status'}
Signed-off-by: Zhang Chen
---
migration/colo.c| 35 +++
qapi
From: zhanghailiang
During the time of VM's running, PVM may dirty some pages, we will transfer
PVM's dirty pages to SVM and store them into SVM's RAM cache at next checkpoint
time. So, the content of SVM's RAM cache will always be same with PVM's memory
after checkpoint.
Instead of flushing all
x27;s failover verdict,
it is still necessary to notify users that we exited COLO mode.
Signed-off-by: zhanghailiang
Signed-off-by: Li Zhijian
Signed-off-by: Zhang Chen
Reviewed-by: Eric Blake
---
migration/colo.c| 20
qapi/migration.json | 35 +
We record the address of the dirty pages that received,
it will help flushing pages that cached into SVM.
Here, it is a trick, we record dirty pages by re-using migration
dirty bitmap. In the later patch, we will start the dirty log
for SVM, just like migration, in this way, we can record both
the
: zhanghailiang
Signed-off-by: Zhang Chen
---
migration/colo.c | 13 +
net/filter-rewriter.c | 40
2 files changed, 53 insertions(+)
diff --git a/migration/colo.c b/migration/colo.c
index 1633aed..6834f3d 100644
--- a/migration/colo.c
+++ b
Filter needs to process the event of checkpoint/failover or
other event passed by COLO frame.
Signed-off-by: zhanghailiang
---
include/net/filter.h | 5 +
net/filter.c | 17 +
net/net.c| 28
3 files changed, 50 insertions(+)
eive all PVM's state.
Signed-off-by: zhanghailiang
Signed-off-by: Li Zhijian
Signed-off-by: Zhang Chen
---
include/exec/ram_addr.h | 1 +
migration/migration.c | 2 +
migration/ram.c | 99 +++--
migration/ram.h | 4 ++
mig
_main() and qemu_save_device_state()
public, and simplify the codes of qemu_save_device_state() by calling the
wrapper qemu_savevm_state_header().
Signed-off-by: zhanghailiang
Signed-off-by: Li Zhijian
Signed-off-by: Zhang Chen
Reviewed-by: Dr. David Alan Gilbert
---
migration/colo.c
Make sure master start block replication after slave's block
replication started.
Besides, we need to activate VM's blocks before goes into
COLO state.
Signed-off-by: zhanghailiang
Signed-off-by: Li Zhijian
Signed-off-by: Zhang Chen
---
migration/colo.c
-by: zhanghailiang
Signed-off-by: Zhang Chen
Reviewed-by: Dr. David Alan Gilbert
---
migration/colo.c | 42 --
migration/migration.c | 4 ++--
2 files changed, 42 insertions(+), 4 deletions(-)
diff --git a/migration/colo.c b/migration/colo.c
index
While do checkpoint, we need to flush all the unhandled packets,
By using the filter notifier mechanism, we can easily to notify
every compare object to do this process, which runs inside
of compare threads as a coroutine.
Signed-off-by: zhanghailiang
Signed-off-by: Zhang Chen
---
include
After a net connection is closed, we didn't clear its releated resources
in connection_track_table, which will lead to memory leak.
Let't track the state of net connection, if it is closed, its related
resources will be cleared up.
Signed-off-by: zhanghailiang
Signed-off-by:
It's a good idea to use notifier to notify COLO frame of
inconsistent packets comparing.
Signed-off-by: Zhang Chen
Signed-off-by: zhanghailiang
---
net/colo-compare.c | 32 +---
net/colo-compare.h | 2 ++
2 files changed, 27 insertions(+), 7 deletions(-)
ble_colo" var" in patch
06/16.
V3:
- Address community comments from V2.
- Rebase on upstream codes.
- Fix several bugs.
- Splite shared disk part to indepentent patch set.
- Optimize codes.
Zhang Chen (10):
filter-rewriter: fix memory leak for connection in
connection_track_t
: zhanghailiang
Signed-off-by: Zhang Chen
---
include/migration/colo.h | 5 ++--
migration/Makefile.objs | 2 +-
migration/colo-comm.c| 76
migration/colo.c | 13 -
migration/migration.c| 23 ++-
migration/savevm.c
On Tue, Feb 6, 2018 at 5:53 PM, Markus Armbruster wrote:
> Zhang Chen writes:
>
> > On Tue, Feb 6, 2018 at 3:27 PM, Markus Armbruster
> wrote:
> >
> >> Zhang Chen writes:
> >>
> >> > On Sat, Feb 3, 2018 at 3:49 PM, Markus Armbruster
> w
On Tue, Feb 6, 2018 at 3:27 PM, Markus Armbruster wrote:
> Zhang Chen writes:
>
> > On Sat, Feb 3, 2018 at 3:49 PM, Markus Armbruster
> wrote:
> >
> >> Zhang Chen writes:
> >>
> >> > From: zhanghailiang
> >> >
> >&
On Sat, Feb 3, 2018 at 3:49 PM, Markus Armbruster wrote:
> Zhang Chen writes:
>
> > From: zhanghailiang
> >
> > If some errors happen during VM's COLO FT stage, it's important to
> > notify the users of this event. Together with 'x-colo-lost-
Hi~ All~
No news for a long time, this is the last part to make COLO running in
upstream.
Any comments for this series?
Thanks
Zhang Chen
On Fri, Jan 19, 2018 at 1:44 PM, Zhang Chen wrote:
> From: Zhang Chen
>
> Hi~
>
> COLO Frame, block replication and COLO proxy(colo-compar
From: zhanghailiang
COLO thread may sleep at qemu_sem_wait(&s->colo_checkpoint_sem),
while failover works begin, It's better to wakeup it to quick
the process.
Signed-off-by: zhanghailiang
---
migration/colo.c | 8
1 file changed, 8 insertions(+)
diff --git a/migration/colo.c b/migra
From: zhanghailiang
Notify all net filters about the checkpoint and failover event.
Signed-off-by: zhanghailiang
---
migration/colo.c | 12
1 file changed, 12 insertions(+)
diff --git a/migration/colo.c b/migration/colo.c
index 9eab4a3..10bc80c 100644
--- a/migration/colo.c
+++ b
_main() and qemu_save_device_state()
public, and simplify the codes of qemu_save_device_state() by calling the
wrapper qemu_savevm_state_header().
Signed-off-by: zhanghailiang
Signed-off-by: Li Zhijian
Signed-off-by: Zhang Chen
Reviewed-by: Dr. David Alan Gilbert
---
migration/colo.c
Filter needs to process the event of checkpoint/failover or
other event passed by COLO frame.
Signed-off-by: zhanghailiang
---
include/net/filter.h | 5 +
net/filter.c | 17 +
net/net.c| 28
3 files changed, 50 insertions(+)
x27;s failover verdict,
it is still necessary to notify users that we exited COLO mode.
Signed-off-by: zhanghailiang
Signed-off-by: Li Zhijian
Signed-off-by: Zhang Chen
Reviewed-by: Eric Blake
---
migration/colo.c| 19 +++
qapi/migration.json | 35
From: zhanghailiang
Don't need to flush all VM's ram from cache, only
flush the dirty pages since last checkpoint
Signed-off-by: Li Zhijian
Signed-off-by: Zhang Chen
Signed-off-by: zhanghailiang
---
migration/ram.c | 12
1 file changed, 12 insertions(+)
diff --git a
From: zhanghailiang
During the time of VM's running, PVM may dirty some pages, we will transfer
PVM's dirty pages to SVM and store them into SVM's RAM cache at next checkpoint
time. So, the content of SVM's RAM cache will always be same with PVM's memory
after checkpoint.
Instead of flushing all
eive all PVM's state.
Signed-off-by: zhanghailiang
Signed-off-by: Li Zhijian
Signed-off-by: Zhang Chen
---
include/exec/ram_addr.h | 1 +
migration/migration.c | 2 +
migration/ram.c | 97 +++--
migration/ram.h | 4 ++
mig
: zhanghailiang
Signed-off-by: Zhang Chen
---
migration/colo.c | 13 +
net/filter-rewriter.c | 40
2 files changed, 53 insertions(+)
diff --git a/migration/colo.c b/migration/colo.c
index a931ff2..9eab4a3 100644
--- a/migration/colo.c
+++ b
We record the address of the dirty pages that received,
it will help flushing pages that cached into SVM.
Here, it is a trick, we record dirty pages by re-using migration
dirty bitmap. In the later patch, we will start the dirty log
for SVM, just like migration, in this way, we can record both
the
-by: zhanghailiang
Signed-off-by: Zhang Chen
Reviewed-by: Dr. David Alan Gilbert
---
migration/colo.c | 42 --
migration/migration.c | 4 ++--
2 files changed, 42 insertions(+), 4 deletions(-)
diff --git a/migration/colo.c b/migration/colo.c
index
Make sure master start block replication after slave's block
replication started.
Besides, we need to activate VM's blocks before goes into
COLO state.
Signed-off-by: zhanghailiang
Signed-off-by: Li Zhijian
Signed-off-by: Zhang Chen
---
migration/colo.c
From: zhanghailiang
We need to know if migration is going into COLO state for
incoming side before start normal migration.
Instead by using the VMStateDescription to send colo_state
from source side to destination side, we use MIG_CMD_ENABLE_COLO
to indicate whether COLO is enabled or not.
Sign
It's a good idea to use notifier to notify COLO frame of
inconsistent packets comparing.
Signed-off-by: Zhang Chen
Signed-off-by: zhanghailiang
---
net/colo-compare.c | 32 +---
net/colo-compare.h | 2 ++
2 files changed, 27 insertions(+), 7 deletions(-)
After a net connection is closed, we didn't clear its releated resources
in connection_track_table, which will lead to memory leak.
Let't track the state of net connection, if it is closed, its related
resources will be cleared up.
Signed-off-by: zhanghailiang
Signed-off-by:
While do checkpoint, we need to flush all the unhandled packets,
By using the filter notifier mechanism, we can easily to notify
every compare object to do this process, which runs inside
of compare threads as a coroutine.
Signed-off-by: zhanghailiang
Signed-off-by: Zhang Chen
---
include
From: Zhang Chen
Hi~
COLO Frame, block replication and COLO proxy(colo-compare,filter-mirror,
filter-redirector,filter-rewriter) have been exist in qemu
for long time, it's time to integrate these three parts to make COLO really
works.
In this series, we have some optimizations for COLO
On Thu, Jan 4, 2018 at 5:10 PM, Eric Blake wrote:
> On 01/04/2018 12:01 AM, Zhang Chen wrote:
> > From: zhanghailiang
> >
> > If some errors happen during VM's COLO FT stage, it's important to
> > notify the users of this event. Together with 'x_colo
From: zhanghailiang
COLO thread may sleep at qemu_sem_wait(&s->colo_checkpoint_sem),
while failover works begin, It's better to wakeup it to quick
the process.
Signed-off-by: zhanghailiang
---
migration/colo.c | 8
1 file changed, 8 insertions(+)
diff --git a/migration/colo.c b/migra
Signed-off-by: zhanghailiang
Signed-off-by: Zhang Chen
---
migration/colo.c | 13 +
net/filter-rewriter.c | 40
2 files changed, 53 insertions(+)
diff --git a/migration/colo.c b/migration/colo.c
index a931ff2..9eab4a3 100644
--- a/migration
From: zhanghailiang
Don't need to flush all VM's ram from cache, only
flush the dirty pages since last checkpoint
Cc: Juan Quintela
Signed-off-by: Li Zhijian
Signed-off-by: Zhang Chen
Signed-off-by: zhanghailiang
---
migration/ram.c | 12
1 file changed, 12 insertion
From: zhanghailiang
Notify all net filters about the checkpoint and failover event.
Cc: Jason Wang
Signed-off-by: zhanghailiang
---
migration/colo.c | 12
1 file changed, 12 insertions(+)
diff --git a/migration/colo.c b/migration/colo.c
index 9eab4a3..10bc80c 100644
--- a/migrat
x27;s failover verdict,
it is still necessary to notify users that we exited COLO mode.
Cc: Markus Armbruster
Cc: Michael Roth
Signed-off-by: zhanghailiang
Signed-off-by: Li Zhijian
Signed-off-by: Zhang Chen
Reviewed-by: Eric Blake
---
migration/colo.c| 19 ++
We record the address of the dirty pages that received,
it will help flushing pages that cached into SVM.
Here, it is a trick, we record dirty pages by re-using migration
dirty bitmap. In the later patch, we will start the dirty log
for SVM, just like migration, in this way, we can record both
the
From: zhanghailiang
During the time of VM's running, PVM may dirty some pages, we will transfer
PVM's dirty pages to SVM and store them into SVM's RAM cache at next checkpoint
time. So, the content of SVM's RAM cache will always be same with PVM's memory
after checkpoint.
Instead of flushing all
Filter needs to process the event of checkpoint/failover or
other event passed by COLO frame.
Cc: Jason Wang
Signed-off-by: zhanghailiang
---
include/net/filter.h | 5 +
net/filter.c | 17 +
net/net.c| 28
3 files changed, 50
Wang
Signed-off-by: zhanghailiang
Signed-off-by: Zhang Chen
Reviewed-by: Dr. David Alan Gilbert
---
migration/colo.c | 42 --
migration/migration.c | 4 ++--
2 files changed, 42 insertions(+), 4 deletions(-)
diff --git a/migration/colo.c b/migration
From: zhanghailiang
We need to know if migration is going into COLO state for
incoming side before start normal migration.
Instead by using the VMStateDescription to send colo_state
from source side to destination side, we use MIG_CMD_ENABLE_COLO
to indicate whether COLO is enabled or not.
Sign
_main() and qemu_save_device_state()
public, and simplify the codes of qemu_save_device_state() by calling the
wrapper qemu_savevm_state_header().
Cc: Juan Quintela
Signed-off-by: zhanghailiang
Signed-off-by: Li Zhijian
Signed-off-by: Zhang Chen
Reviewed-by: Dr. David Alan Gilbert
---
migration/c
eive all PVM's state.
Cc: Dr. David Alan Gilbert
Signed-off-by: zhanghailiang
Signed-off-by: Li Zhijian
Signed-off-by: Zhang Chen
---
include/exec/ram_addr.h | 1 +
migration/migration.c | 2 +
migration/ram.c | 97 +++--
mig
While do checkpoint, we need to flush all the unhandled packets,
By using the filter notifier mechanism, we can easily to notify
every compare object to do this process, which runs inside
of compare threads as a coroutine.
Cc: Jason Wang
Signed-off-by: zhanghailiang
Signed-off-by: Zhang Chen
It's a good idea to use notifier to notify COLO frame of
inconsistent packets comparing.
Cc: Jason Wang
Signed-off-by: Zhang Chen
Signed-off-by: zhanghailiang
---
net/colo-compare.c | 32 +---
net/colo-compare.h | 2 ++
2 files changed, 27 insertions(
Make sure master start block replication after slave's block
replication started.
Besides, we need to activate VM's blocks before goes into
COLO state.
Signed-off-by: zhanghailiang
Signed-off-by: Li Zhijian
Signed-off-by: Zhang Chen
Cc: Stefan Hajnoczi
Cc: Kevin Wolf
Cc: Max Reit
After a net connection is closed, we didn't clear its releated resources
in connection_track_table, which will lead to memory leak.
Let't track the state of net connection, if it is closed, its related
resources will be cleared up.
Signed-off-by: zhanghailiang
Signed-off-by:
plite shared disk part to indepentent patch set.
- Optimize codes.
Zhang Chen (8):
filter-rewriter: fix memory leak for connection in
connection_track_table
colo-compare: implement the process of checkpoint
colo-compare: use notifier to notify packets comparing result
COLO: integrate
|6|7|8|9|0| spkt1
> ---|\|
>| \offset |
> -v-v
> | header |4|5|6|7|8|9|0| ppkt2
>
>
> In this way, the performance can reach native 20% in my multiple
> tests.
On Fri, Dec 22, 2017 at 5:28 PM, Mao Zhongyi
wrote:
> Modified the function colo_packet_compare_common to prepare for the
> tcp packet comparison in the next patch.
>
> Cc: Zhang Chen
> Cc: Li Zhijian
> Cc: Jason Wang
>
> Signed-off-by: Mao Zhongyi
> Signed-off-b
On Wed, Dec 13, 2017 at 6:25 AM, Mao Zhongyi
wrote:
> Hi, Chen
>
> On 12/12/2017 11:21 PM, Zhang Chen wrote:
>
>>
>>
>> On Wed, Dec 6, 2017 at 5:57 PM, Mao Zhongyi > <mailto:maozy.f...@cn.fujitsu.com>> wrote:
>>
>> The primary and secon
On Wed, Dec 13, 2017 at 5:32 AM, Mao Zhongyi
wrote:
> Hi, Chen
>
> On 12/12/2017 11:19 PM, Zhang Chen wrote:
>
>> This patch should be moved behind the patch with payload comparition.
>>
>
> Well, it is not impossible, but I think it is better to be here. Because
On Wed, Dec 6, 2017 at 5:57 PM, Mao Zhongyi
wrote:
> Cc: Zhang Chen
> Cc: Li Zhijian
> Cc: Jason Wang
>
> Signed-off-by: Mao Zhongyi
> ---
> net/colo-compare.c | 16
> net/colo.c | 1 +
> net/colo.h | 1 +
> net/trace-events
:
>
> | header |1|2|3| ppkt1
> -|-|
> | |
> -v-v--
> | header |1|2|3|4|5|6|7|8|9|0| spkt1
> ---|\|
>| \offset |
> -----v-v
> | header |4|5|6|7|8|9|0| ppkt2
>
On Wed, Dec 6, 2017 at 5:57 PM, Mao Zhongyi
wrote:
> Modified the function colo_packet_compare_common to prepare for the
> tcp packet comparison in the next patch.
>
> Cc: Zhang Chen
> Cc: Li Zhijian
> Cc: Jason Wang
>
> Signed-off-by: Mao Zhongyi
> ---
This patch should be moved behind the patch with payload comparition.
People needs look your changes to understand why we need this patch.
Thanks
Zhang Chen
On Wed, Dec 6, 2017 at 9:57 AM, Mao Zhongyi
wrote:
> The th_off field specifies the size of the TCP header, if th_off > 5,
>
On Mon, Dec 4, 2017 at 3:32 AM, Mao Zhongyi
wrote:
>
>
> On 12/04/2017 09:41 AM, Zhang Chen wrote:
>
>>
>>
>> On Tue, Nov 28, 2017 at 8:04 PM, Mao Zhongyi > <mailto:maozy.f...@cn.fujitsu.com>> wrote:
>>
>> The primary and secondary
On Tue, Nov 28, 2017 at 8:04 PM, Mao Zhongyi
wrote:
> Cc: Zhang Chen
> Cc: Li Zhijian
> Cc: Jason Wang
>
> Signed-off-by: Mao Zhongyi
> ---
> net/colo-compare.c | 16
> net/colo.c | 1 +
> net/colo.h | 1 +
> net/trace-event
| header |1|2|3| ppkt1
> -|-|
> | |
> -v-v--
> | header |1|2|3|4|5|6|7|8|9|0| spkt1
> ---|\|
>| \offset |
> -v-v
> | header |4|5|6|7|8|9|0| ppkt2
> ---
On 08/29/2017 03:22 PM, Wang yong wrote:
From: Wang Yong
Update colo-proxy.txt,add IOThread configuration.
Later we have to configure IOThread,if not COLO can not work.
Signed-off-by: Wang Yong
Signed-off-by: Wang Guang
Reviewed-by: Zhang Chen
---
docs/colo-proxy.txt | 3 ++-
1
g_main_loop in
the IOThread.then the packets from the primary and the secondary
are processed in the IOThread.
Finally remove the colo-compare thread using the IOThread instead.
Signed-off-by: Wang Yong
Signed-off-by: Wang Guang
Reviewed-by: Zhang Chen
Hi~ Jason.
Have any comments to this series
When network is busy, some tcp options(like sack) will unpredictable
occur in primary side or secondary side. it will make packet size
not same, but the two packet's payload is identical. colo just
care about packet payload, so we skip the option field.
Signed-off-by: Zhang Chen
---
net
The packet_enqueue() use g_queue_push_tail() to
enqueue net packet, so it is more efficent way use
g_queue_pop_head() to get packet for compare.
That will improve the success rate of comparison.
In my test the performance of ftp put 1000M file
will increase 10%
Signed-off-by: Zhang Chen
---
net
Signed-off-by: Zhang Chen
---
net/colo-compare.c | 59 --
1 file changed, 31 insertions(+), 28 deletions(-)
diff --git a/net/colo-compare.c b/net/colo-compare.c
index c31adc6..d5fcfea 100644
--- a/net/colo-compare.c
+++ b/net/colo-compare.c
p2.
- Move the "checkpoint_time_ms" to CompareState,
in order to aviod multi colo-compare instance conflict.
- Add "TODO comments" for reset s->checkpoint_time_ms.
- Add a new patch fix comments and scheme.
V2:
- Rename p2's subject.
Zhang Chen (3):
On 09/05/2017 10:11 AM, Dou Liyang wrote:
Hi Chen,
At 09/04/2017 02:14 PM, Zhang Chen wrote:
When network is busy, some tcp options(like sack) will unpredictable
occur in primary side or secondary side. it will make packet size
not same, but the two packet's payload is identical. colo
When network is busy, some tcp options(like sack) will unpredictable
occur in primary side or secondary side. it will make packet size
not same, but the two packet's payload is identical. colo just
care about packet payload, so we skip the option field.
Signed-off-by: Zhang Chen
---
net
Signed-off-by: Zhang Chen
---
net/colo-compare.c | 59 --
1 file changed, 31 insertions(+), 28 deletions(-)
diff --git a/net/colo-compare.c b/net/colo-compare.c
index 8dd76e4..c9a7174 100644
--- a/net/colo-compare.c
+++ b/net/colo-compare.c
s" to CompareState,
in order to aviod multi colo-compare instance conflict.
- Add "TODO comments" for reset s->checkpoint_time_ms.
- Add a new patch fix comments and scheme.
V2:
- Rename p2's subject.
Zhang Chen (3):
net/colo-compare.c: Optimize unpredictable tcp o
The packet_enqueue() use g_queue_push_tail() to
enqueue net packet, so it is more efficent way use
g_queue_pop_head() to get packet for compare.
That will improve the success rate of comparison.
In my test the performance of ftp put 1000M file
will increase 10%
Signed-off-by: Zhang Chen
---
net
901 - 1000 of 1727 matches
Mail list logo
