Global Services no longer shadow locals?

[George_Ellis]

I am pretty sure of the answer, because Phoneboy mentioned it and the evidence is there from the migration to R80.10. 

The question is: are Global Services finally usable? 

In the past, a global service and a local service for the same port could shadow each other and the compiler would not catch it in verification.  If that has been truly corrected, that means we could start using global service objects.  This will simplify our automation efforts.

TIA

[PhoneBoy]

I think what I said was that you can't modify the default services anymore.

But still, it's a fair question.

[Ofer_Barzvi]

In pre-R80 releases, 'g' was added as a prefix to global services and therefore the rule-hiding-rule logic considered the global and local services as different services as they are with different names.

However, in R80.x releases, the 'g' prefix isn't part of the global service name and therefore you will get a verification failure in case of conflict between global and local rules with same service.

[George_Ellis]

Our Diamond engineer just tested a port 80 global vs a port 80 local and they verified without detecting they were  shadowing.

[PhoneBoy]

I assume an SR was opened on this, correct?

Because that doesn't sound like correct behavior from what Ofer said above.

[George_Ellis]

That is how Diamond was involved, an SR.  I can give you the id privately.