Re: Repository services and budget

[Marcel Schneider <schneider@switch.ch>]

On Monday, 13 Jan 1997, "Rick H. Wesson" writes:
> On Jan 13,  2:23pm, Dave Crocker wrote:
>> Subject: Re: Repository services and budget
>>
>> >
>> >> >You are not going to be able to do this under a million dollars and
>> >> >guarantee round the clock service to the internet community. Registrars
>> >>
>> >>         Please provide a detailed budget and an explanation for the
>> >> functions provided by each line item.
>>
>> 	Yes, it is necessary.  A detailed budget produced a result that was
>> 1/2-1/4 the estimate cited here.  If one is going to assert that a budget
>> needs to be 2-4 times larger, one should provide the basis for the
>> estimate.  In the absence of such detail, the estimate is merely a handwave.
>>
> Dave,

> Your estimate was for a cripled set of linux boxen, and provides little
> insentive for those that run the thing for any type of excellance. If
> business are going to invest in running a Registry, it had better dam
> well work 24 hours a day and 7 daze a week!

> Ongoing
> ===============================================
> 3 software developers 	80K each/yr	240,000
> 3 T-1 Connections	12K each/yr	 36,000
> 1 Manager/Director			120,000
> 3 Rotateing 24/7
>     NOC support ppl	35K each/yr	105,000
> Travel					 80,000
> Telephone				 35,000
> Legal Team				125,000

> Startup
> ==============================================
> Routers					 25,000
> Database Software			 45,000
> PC/Unix workstations    10@5K eack	 50,000
> Database Servers 	3@30K each	 90,000
> RAID disk Storage 			 30,000
> Online Backup.				 30,000

> 			Sub Total      1,011,000

> If you add public key crypto and/or distribute whitepages info add.
> RSA/IDEA License			 45,000
> 3 dispute resolution staff		150,000
> 2 developers (pgp keyring)		160,000
> 1 server to publish the keyring		 20,000

> 			Sub Total	375,000

> 			Grand Total   1,386,000

Having set up a registry last year (TLD's CH and LI) I can fully
support above figures. It is even more expensive since accounting
has been omitted and this cannot be neglected (shipping thousends
of invoices and trying to file payments and trying to trace
payments etc.). Add another person for accounting and approx.
USD 100k for software. The legal costs are closely related to
the policy established (liberal/strict, involvment/non involvment
in disputes).

The number of staff required depends on number of registrations
per time.

A rough figure might be 1'000 _new_ registrations per month can
be managed by approx. 3 full time persons. Note that an overhead
of approx. factor 4.5 is to be calculated for all other stuff 
(deletions, modifications, rejects of new applications, communication 
via mail/fax/phone, accounting) per _successful_ registration.
OK, the figure 4.5 is only what we have experienced last year
and strongly depends how easily applicants understand what they
are required to do :). As a rule, nobody RTFM.

The main tasks to be solved software-wise are

a) automate registration (implement policy, feed WHOIS)
b) modify existing WHOIS to your needs (we had to include
   billing contact, active/inactive, person/org and other
   new tags, adopt it to Solaris)
c) automate billing (we automatically print invoices of the
   previous day but still have to manually ship these)
d) automate filing of payments
e) send fax/e-mail with confirmation message for registered
   names and reasons for rejects automatically
f) implement trouble ticket system.

In addition to above costs you will have to print documents, will
need a lot of stamps, invoices and envelopes and have expenditures
for service contracts and charges by the bank(s). When you have to
care for oversea connectivity (as we poor Europeans have ;-))
then expenditures soar accordingly.  

> The CORE database is to be used by BUSINESSES that are to be providing
> COMPETITIVE services against NSI. Why should it not be done as robustly
> as possable? Afterall these businesses are betting their livelyhood on the
> reliblity of this system.

> It does puzzle me as to why the IAHC does not want to discuss the functionality
> of the CORE repository, as that functionality defines what and how services
> are sold by the registrys that win the lottery.

> I run a business and need much more than a gut feeling about the viability of
> a domain registry to determin if it is worthwhile to persue submitting a,
> yet to be defined, application to run a registry.

> I would like to add that if the CORE repository is run at a fixed cost,
> then there is no privisions for growth. That is the main reason
> NSI started chargeing for domain names in the first place, they couldn't
> handle the growth.

> If all this is to be determined by the CORE I could only say that for the first
> time "on the wire" protocols will be defined by a small closed group and not
> by a Working Group process. Is this wize?

> -Rick (waiving hand)


> -- 
> Rick H. Wesson


Marcel Schneider@SWITCH