Hi Lorenzo, Ilya,

Thanks for the patch and review!

On 6/30/26 3:53 PM, Ilya Maximets wrote:
> On 6/30/26 3:17 PM, Lorenzo Bianconi wrote:
>> Always use sb_table_set_req_mon_condition macro in update_sb_monitors
>> routine even for optional tables (e.g. advertised_mac_binding or
>> advertised_route) since OvS IDL layer will take care of creating proper
>> monitor requests according to the real db schema.
> 
> This will only be true after we move submodule to the current ovs/main.
> Before that the change is not safe to apply.
> 

Is there any reason why we're not bumping the submodule in this patch?

>> This patch fixes the conditional monitor support for optional tables
>> since sb_table_set_opt_mon_condition macro will not properly enable
>> conditional monitoring until we run update_sb_monitors() in the
>> ovn-controller main loop. As result we will always receive notifications
>> for all the columns of optional tables (even if monitor-all is not enabled)
>> until we modify a monitor conditions.
>>
>> Reported-at: https://issues.redhat.com/browse/FDP-3114
>> Signed-off-by: Lorenzo Bianconi <[email protected]>
>> ---
>>  controller/ovn-controller.c | 41 +++++++++++++++----------------------
>>  1 file changed, 16 insertions(+), 25 deletions(-)
>>
>> diff --git a/controller/ovn-controller.c b/controller/ovn-controller.c
>> index 7e6c9e69a..06736395b 100644
>> --- a/controller/ovn-controller.c
>> +++ b/controller/ovn-controller.c
>> @@ -215,15 +215,7 @@ static char *get_file_system_id(void)
>>      free(filename);
>>      return ret;
>>  }
>> -/* Only set monitor conditions on tables that are available in the
>> - * server schema.
>> - */
>> -#define sb_table_set_opt_mon_condition(idl, table, cond) \
>> -    (sbrec_server_has_##table##_table(idl)               \
>> -     ? sbrec_##table##_set_condition(idl, cond)          \
>> -     : 0)
>>  
>> -/* Assume the table exists in the server schema and set its condition. */
>>  #define sb_table_set_req_mon_condition(idl, table, cond) \
>>      sbrec_##table##_set_condition(idl, cond)
> 
> Should this also be removed?
> 

Looks like it, yes.

> Best regards, Ilya Maximets.
> 

Regards,
Dumitru

_______________________________________________
dev mailing list
[email protected]
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Reply via email to