Therefore, personally, I think it's time for us to start thinking how to set up SLA's between LIRs and RIPE NCC: let's define strict time boundaries for the first and later hostmaster answers to an IP/AS request. It doesn't matter whether it is 5, 7 or 14 days - but let's define them and abide by them. Also, an escalation procedure within RIPE NCC is needed (who to contact should a request get not processed - e.g. Registration services manager and higher), determine what happens should RIPE NCC fail to process a request within specified time (e.g. deduce XXX EUR from next year contribution fee for each day of delay).

That doesn't clear the problem. The lack of resources don't vanish if issues are escalated to managers and such. Or if they do vanish we have a problem with work-moral within hostmasters. There is a risk with this suggestion that service-time (SLA) would be different for different LIR's, and the LIR whose hostmaster yells out load will get serviced first. If we reduce the annual contribution fee when a request is not served within the SLA, we are only making things worse. RIPE NCC will then have even less money to correct the problem. Maybe we should try to give somekind of carrot to NCC. If IP/ASN request is served less than 5 working days give 100EUR extra contribution, if it is served in less than 2 working days give 200EUR. This is also an equal method, you cannot use money to buy SLA's. Jorma ---------------------------------------------------------- jorma.mellin@teliafi.net Development Manager ; CCIE#4185 Telia Finland Inc, Network Services