Re: [dns-wg] Progress with DNS Quality, Also: Lameness
"Daniel" == Daniel Karrenberg <daniel.karrenberg@ripe.net> writes:
Daniel> How about your friendly RIR providing an extra service of Daniel> checking your reverse DNS tree for a small additional fee Daniel> that covers the people resources needed to get the reports Daniel> to you and follow up on them. I agree that someone could/should offer a service selling police reports. And related stuff such as audits and logs: "Yes, your name servers were working just fine between 14:40 and 16:00 on June 1st 2002". However I am very uncomfortable with an RIR providing this kind of service, even though they're more likely to be respected as protocol police. There are a number of reasons for this. First of all I'm not sure if services like this fall within the remit of what an RIR is supposed to do. Secondly an RIR is a monopoly and should be *very* careful about extending into new service areas. The likes of the EU anti-competition folks could get very excited about that. Thirdly, this the kind of service that could be offered by others: ISPs, registrars, DNS consulting companies and so on. Many of them will already be paying fees to the RIR. An RIR shouldn't be competing with its customers -- why bite the hand that feeds it? -- or distorting the market. Fourthly, if an RIR diversifies into offering these peripheral services, there is a danger that it will lose sight of its core registry function. I do like the idea of RIRs having top-up charging for additional services. This is a good way to make those services self-financing and be seen to not be a drain on the RIR's core funding. And if that keeps the annual fees down, even better. However this is a discussion that should be going on somewhere else: the new NCC services WG perhaps? Finding out who'd be willing to pay for lameness checks would be worthwhile. However I suspect it won't tell us anything we don't already know. Clueful people might pay. But in general they don't have lame servers. The clueless won't pay because they don't know or care about lameness. They'll be the ones that have the lame delegations. The challenge will always be how to reach out to these people. There's an education problem here. I think the WG could come up with a definition of lameness and a business justification why it's a bad thing. It would be great if registrars & registries could pick this up, point their customers at checking tools and persuade them to regularly use those tools. This is do-able IMO.
We have been developing such a tool. Currently it has about 40 million zones loaded in a SQL database. We have been trying to get ccTLD operators to let us slave or FTP their zones as well. Some let us, many others dont. Access to the tool is based on setting up an account. If you are interested send email to data AT chagres DOT net with your UID, IP addr, Company Affiliation, Statement that you will NOT use for SPAM or HACKING. We will send your password and URL back to you. Today the tool lets you type in ripe.net it returns the list of NS's that end with ripe.net and a count of the zones in the DB for each NS. You can then get a list of zones, or do a LAME check. You can also create a list thats the DIFF between two different NS"s. Say you have NS1.EXAMPLE.COM and NS2.EXAMPLE.COM NS1 has 300 zones and NS2 has 281 zones. You know that they should be the same. You can diff the NS1 and NS2 lists to see whats missing. cheers john brown chagres technologies, inc
-----Original Message----- From: dns-wg-admin@ripe.net [mailto:dns-wg-admin@ripe.net] On Behalf Of Jim Reid Sent: Thursday, May 22, 2003 3:05 AM To: RIPE DNS Working Group Subject: Re: [dns-wg] Progress with DNS Quality, Also: Lameness
"Daniel" == Daniel Karrenberg <daniel.karrenberg@ripe.net> writes:
Daniel> How about your friendly RIR providing an extra service of Daniel> checking your reverse DNS tree for a small additional fee Daniel> that covers the people resources needed to get the reports Daniel> to you and follow up on them.
I agree that someone could/should offer a service selling police reports. And related stuff such as audits and logs: "Yes, your name servers were working just fine between 14:40 and 16:00 on June 1st 2002". However I am very uncomfortable with an RIR providing this kind of service, even though they're more likely to be respected as protocol police.
There are a number of reasons for this. First of all I'm not sure if services like this fall within the remit of what an RIR is supposed to do. Secondly an RIR is a monopoly and should be *very* careful about extending into new service areas. The likes of the EU anti-competition folks could get very excited about that. Thirdly, this the kind of service that could be offered by others: ISPs, registrars, DNS consulting companies and so on. Many of them will already be paying fees to the RIR. An RIR shouldn't be competing with its customers -- why bite the hand that feeds it? -- or distorting the market. Fourthly, if an RIR diversifies into offering these peripheral services, there is a danger that it will lose sight of its core registry function.
I do like the idea of RIRs having top-up charging for additional services. This is a good way to make those services self-financing and be seen to not be a drain on the RIR's core funding. And if that keeps the annual fees down, even better. However this is a discussion that should be going on somewhere else: the new NCC services WG perhaps?
Finding out who'd be willing to pay for lameness checks would be worthwhile. However I suspect it won't tell us anything we don't already know. Clueful people might pay. But in general they don't have lame servers. The clueless won't pay because they don't know or care about lameness. They'll be the ones that have the lame delegations. The challenge will always be how to reach out to these people.
There's an education problem here. I think the WG could come up with a definition of lameness and a business justification why it's a bad thing. It would be great if registrars & registries could pick this up, point their customers at checking tools and persuade them to regularly use those tools. This is do-able IMO.
participants (2)
-
Jim Reid -
john brown