methods of measurements
Dear all, in section 2.5 Appendix D, Table D-2, Performance Specifications of VeriSigns .net application, VeriSign specifies 50 milliseconds as maximum for an add/create command and 25 milliseconds for a check command. Furthermore in the ICANN monthly reports (see e.g. http://www.icann.org/tlds/monthly-reports/com-net/verisign-200409.pdf, page 3) VeriSign reports 10 milliseconds as an average value for add/create and check commands. In comparison to other applications for .net these values differ in at least on decimal power. (I do not want to discuss here whether such a race of milliseconds make sense to registrars or registrants and leave this discussion to the evaluators.) On the other hand the registrar department of CORE has made a long term statistics over some month. Our results reports for VeriSign an average of 647 ms for add/create and 426 ms for a check command. This includes a round trip time of 90 ms. It was measured the time from sending the command until receiving an answer from the registry, not including the processing of the commands. description add/create check ---------------------------------- ---------- ---------- proposed SLA by VeriSign 50 ms 25 ms published ICANN monthly stats 10 ms 10 ms long term measurement by registrar 557 ms 336 ms As I am involved in the CORE++ bid, I want to make it very clear that I do not try to denigrate or mistrust any other proposal. I take every applicant as reputable. But I have the fear that applicants use several methods of measurements and therefore such huge differences appear. Perhaps VeriSign can disclose their methods and explain their numbers in more detail. I ask for VeriSign, because the other applicants have numbers in the same magnitude. Best regards, Elmar --
_____________________________________________________________________
| |
| knipp | Knipp Medien und Kommunikation GmbH
------- Technologiepark
Martin-Schmeisser-Weg 9
44227 Dortmund
Germany
Fon: +49 231 9703-0
Elmar.Knipp@xxxxxxxx Fax: +49 231 9703-200
|