Re: Allow parallel plan for referential integrity checks?

On 4/14/22 14:25, Frédéric Yhuel wrote:
> 
> 
> On 3/19/22 01:57, Imseih (AWS), Sami wrote:
>> I looked at your patch and it's a good idea to make foreign key 
>> validation
>> use parallel query on large relations.
>>
>> It would be valuable to add logging to ensure that the ActiveSnapshot 
>> and TransactionSnapshot
>> is the same for the leader and the workers. This logging could be 
>> tested in the TAP test.
>>
>> Also, inside RI_Initial_Check you may want to set max_parallel_workers to
>> max_parallel_maintenance_workers.
>>
>> Currently the work_mem is set to maintenance_work_mem. This will also 
>> require
>> a doc change to call out.
>>
>> /*
>>       * Temporarily increase work_mem so that the check query can be 
>> executed
>>       * more efficiently.  It seems okay to do this because the query 
>> is simple
>>       * enough to not use a multiple of work_mem, and one typically 
>> would not
>>       * have many large foreign-key validations happening 
>> concurrently.  So
>>       * this seems to meet the criteria for being considered a 
>> "maintenance"
>>       * operation, and accordingly we use maintenance_work_mem.  
>> However, we
>>
> 
> Hello Sami,
> 
> Thank you for your review!
> 
> I will try to do as you say, but it will take time, since my daily job 
> as database consultant takes most of my time and energy.
> 

Hi,

As suggested by Jacob, here is a quick message to say that I didn't find 
enough time to work further on this patch, but I didn't completely 
forget it either. I moved it to the next commitfest. Hopefully I will 
find enough time and motivation in the coming months :-)

Best regards,
Frédéric