[Bizgres-general] Statement Queuing take II - Resource Scheduling

[Mark Kirkwood]

Jim C. Nasby wrote:
> On Thu, Jun 29, 2006 at 05:34:26PM +1200, Mark Kirkwood wrote:
>   
>>
>> I wondering about the definition of 'lowest role in the tree':
>>
>>  If it is 'the role with the lowest resource limits', then this is 
>> potentially not defined when there is more than one resource limit type 
>> - e.g. (inventing some extra limit types as we go) suppose  role 'A' has 
>> cpu limit 90 and memory limit 80  and role 'B' has  them the other way 
>> around, if I am a member of 'A' and 'B' it is not possible to decide 
>> which role to use with this definition.
>>     
>  
> You could take the minimum of all limits...
>   
This idea was running through my head as well - I was hoping there might 
be  a simpler way :-).

>> I'm wondering if there needs to be a check along the lines of 'you can 
>> only be a member of 1 group with resource limits enabled'?
>>
>>     
>
> I'm thinking that for the first pass maybe we should only look at
> whatever role you have active at the time. Or, take the minimum of all
> limits that apply to roles that are directly INHERITed by the current
> role, if SET ROLE hasn't been used. Once we have a better idea of how
> people are using resource limits, I think it'll be a lot easier to
> decide what to do when there's a whole tree of limits to deal with.
>   
Right - The SET ROLE is a bit alarming, as it means users can 'switch 
off' their resource limits in some cases. I guess it means extra care is 
needed during ROLE setup to ensure that tables you wish to  protect with 
resource scheduling cannot be accessed after SET ROLE 
role-with-no-resource-limits. How much of a risk do you think this is?

> Though, if this is the route we take, we should make it very clear in
> the docs that things will change in a future version.
>   
Sure.

Cheers

Mark