DPDK patches and discussions
 help / color / mirror / Atom feed
* [dpdk-dev] Question about bonding port promiscuous
@ 2021-10-14  2:13 Min Hu (Connor)
  2021-10-19 11:52 ` Min Hu (Connor)
  0 siblings, 1 reply; 7+ messages in thread
From: Min Hu (Connor) @ 2021-10-14  2:13 UTC (permalink / raw)
  To: dev; +Cc: Declan Doherty

Hi, Declan Doherty,
	In function "bond_ethdev_promiscuous_disable",
for "ROUND ROBIN","BALANCE","BROADCAST","8023AD" mode, Promiscuous mode
is propagated to all slaves.

While for "ACTIVE_BACKUP", "TLB", "ALB", promiscuous mode is propagated 
only to primary slave.

Why?

The second question, for "ACTIVE_BACKUP", when primary active slave
fails, a different slave becomes active, does the promiscuous mode state
of former primary should be set to the new primary?

	Hope for your reply.
	Best Wishes.

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [dpdk-dev] Question about bonding port promiscuous
  2021-10-14  2:13 [dpdk-dev] Question about bonding port promiscuous Min Hu (Connor)
@ 2021-10-19 11:52 ` Min Hu (Connor)
  2021-10-21 12:58   ` Min Hu (Connor)
  0 siblings, 1 reply; 7+ messages in thread
From: Min Hu (Connor) @ 2021-10-19 11:52 UTC (permalink / raw)
  To: dev; +Cc: Declan Doherty, Ferruh Yigit, Thomas Monjalon

Hi, anyone could give me a reply? thanks.

在 2021/10/14 10:13, Min Hu (Connor) 写道:
> Hi, Declan Doherty,
>      In function "bond_ethdev_promiscuous_disable",
> for "ROUND ROBIN","BALANCE","BROADCAST","8023AD" mode, Promiscuous mode
> is propagated to all slaves.
> 
> While for "ACTIVE_BACKUP", "TLB", "ALB", promiscuous mode is propagated 
> only to primary slave.
> 
> Why?
> 
> The second question, for "ACTIVE_BACKUP", when primary active slave
> fails, a different slave becomes active, does the promiscuous mode state
> of former primary should be set to the new primary?
> 
>      Hope for your reply.
>      Best Wishes.
> .

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [dpdk-dev] Question about bonding port promiscuous
  2021-10-19 11:52 ` Min Hu (Connor)
@ 2021-10-21 12:58   ` Min Hu (Connor)
  2021-10-21 13:39     ` Min Hu (Connor)
  0 siblings, 1 reply; 7+ messages in thread
From: Min Hu (Connor) @ 2021-10-21 12:58 UTC (permalink / raw)
  To: dev; +Cc: Declan Doherty, Ferruh Yigit, Thomas Monjalon

Hi, Ferruh, Thomas,
	any comments?

在 2021/10/19 19:52, Min Hu (Connor) 写道:
> Hi, anyone could give me a reply? thanks.
> 
> 在 2021/10/14 10:13, Min Hu (Connor) 写道:
>> Hi, Declan Doherty,
>>      In function "bond_ethdev_promiscuous_disable",
>> for "ROUND ROBIN","BALANCE","BROADCAST","8023AD" mode, Promiscuous mode
>> is propagated to all slaves.
>>
>> While for "ACTIVE_BACKUP", "TLB", "ALB", promiscuous mode is 
>> propagated only to primary slave.
>>
>> Why?
>>
>> The second question, for "ACTIVE_BACKUP", when primary active slave
>> fails, a different slave becomes active, does the promiscuous mode state
>> of former primary should be set to the new primary?
>>
>>      Hope for your reply.
>>      Best Wishes.
>> .
> .

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [dpdk-dev] Question about bonding port promiscuous
  2021-10-21 12:58   ` Min Hu (Connor)
@ 2021-10-21 13:39     ` Min Hu (Connor)
  0 siblings, 0 replies; 7+ messages in thread
From: Min Hu (Connor) @ 2021-10-21 13:39 UTC (permalink / raw)
  To: dev; +Cc: Declan Doherty, Ferruh Yigit, Thomas Monjalon, David Marchand

Hi, David,
	how about your opinion ?

在 2021/10/21 20:58, Min Hu (Connor) 写道:
> Hi, Ferruh, Thomas,
>      any comments?
> 
> 在 2021/10/19 19:52, Min Hu (Connor) 写道:
>> Hi, anyone could give me a reply? thanks.
>>
>> 在 2021/10/14 10:13, Min Hu (Connor) 写道:
>>> Hi, Declan Doherty,
>>>      In function "bond_ethdev_promiscuous_disable",
>>> for "ROUND ROBIN","BALANCE","BROADCAST","8023AD" mode, Promiscuous mode
>>> is propagated to all slaves.
>>>
>>> While for "ACTIVE_BACKUP", "TLB", "ALB", promiscuous mode is 
>>> propagated only to primary slave.
>>>
>>> Why?
>>>
>>> The second question, for "ACTIVE_BACKUP", when primary active slave
>>> fails, a different slave becomes active, does the promiscuous mode state
>>> of former primary should be set to the new primary?
>>>
>>>      Hope for your reply.
>>>      Best Wishes.
>>> .
>> .
> .

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [dpdk-dev] Question about bonding port promiscuous
  2021-10-27 13:47 ` David Marchand
@ 2021-10-28  2:08   ` Min Hu (Connor)
  0 siblings, 0 replies; 7+ messages in thread
From: Min Hu (Connor) @ 2021-10-28  2:08 UTC (permalink / raw)
  To: David Marchand; +Cc: dev



在 2021/10/27 21:47, David Marchand 写道:
> On Wed, Oct 27, 2021 at 5:21 AM Min Hu (Connor) <humin29@huawei.com> wrote:
>>
>> Hi, David,
>> In function "bond_ethdev_promiscuous_disable",
>> for "ROUND ROBIN","BALANCE","BROADCAST","8023AD" mode, Promiscuous mode
>> is propagated to all slaves.
>>
>> While for "ACTIVE_BACKUP", "TLB", "ALB", promiscuous mode is propagated
>> only to primary slave.
>>
>> Why?
> 
>>From what I remember of the rework I did some time ago, the code was
> already doing it and I did not question it.
> I guess it mimicks what the Linux kernel bonding driver does.
> 
>>
>> The second question, for "ACTIVE_BACKUP", when primary active slave
>> fails, a different slave becomes active, does the promiscuous mode state
>> of former primary should be set to the new primary?
> 
> IOW, setting promisc on the link that becomes primary?
> If it is not done, it sounds like a bug.
OK, I will try to fix it, thanks David.
> 
> 

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [dpdk-dev] Question about bonding port promiscuous
  2021-10-27  3:21 Min Hu (Connor)
@ 2021-10-27 13:47 ` David Marchand
  2021-10-28  2:08   ` Min Hu (Connor)
  0 siblings, 1 reply; 7+ messages in thread
From: David Marchand @ 2021-10-27 13:47 UTC (permalink / raw)
  To: Min Hu (Connor); +Cc: dev

On Wed, Oct 27, 2021 at 5:21 AM Min Hu (Connor) <humin29@huawei.com> wrote:
>
> Hi, David,
> In function "bond_ethdev_promiscuous_disable",
> for "ROUND ROBIN","BALANCE","BROADCAST","8023AD" mode, Promiscuous mode
> is propagated to all slaves.
>
> While for "ACTIVE_BACKUP", "TLB", "ALB", promiscuous mode is propagated
> only to primary slave.
>
> Why?

From what I remember of the rework I did some time ago, the code was
already doing it and I did not question it.
I guess it mimicks what the Linux kernel bonding driver does.

>
> The second question, for "ACTIVE_BACKUP", when primary active slave
> fails, a different slave becomes active, does the promiscuous mode state
> of former primary should be set to the new primary?

IOW, setting promisc on the link that becomes primary?
If it is not done, it sounds like a bug.


-- 
David Marchand


^ permalink raw reply	[flat|nested] 7+ messages in thread

* [dpdk-dev] Question about bonding port promiscuous
@ 2021-10-27  3:21 Min Hu (Connor)
  2021-10-27 13:47 ` David Marchand
  0 siblings, 1 reply; 7+ messages in thread
From: Min Hu (Connor) @ 2021-10-27  3:21 UTC (permalink / raw)
  To: David Marchand; +Cc: dev

Hi, David,
In function "bond_ethdev_promiscuous_disable",
for "ROUND ROBIN","BALANCE","BROADCAST","8023AD" mode, Promiscuous mode
is propagated to all slaves.

While for "ACTIVE_BACKUP", "TLB", "ALB", promiscuous mode is propagated 
only to primary slave.

Why?

The second question, for "ACTIVE_BACKUP", when primary active slave
fails, a different slave becomes active, does the promiscuous mode state
of former primary should be set to the new primary?

      Hope for your reply.
      Best Wishes.
.

^ permalink raw reply	[flat|nested] 7+ messages in thread

end of thread, other threads:[~2021-10-28  2:08 UTC | newest]

Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-10-14  2:13 [dpdk-dev] Question about bonding port promiscuous Min Hu (Connor)
2021-10-19 11:52 ` Min Hu (Connor)
2021-10-21 12:58   ` Min Hu (Connor)
2021-10-21 13:39     ` Min Hu (Connor)
2021-10-27  3:21 Min Hu (Connor)
2021-10-27 13:47 ` David Marchand
2021-10-28  2:08   ` Min Hu (Connor)

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).