Puzzle of the week, win a pie :-)

Written by Roland van Rijswijk in category: General, Users

I’ve created a new validation rate graph based on the statistics gathered on our resolvers. The data is up-to-date until week 43 (this week). The graph shows an interesting trend:

In the graph you can see three lines representing three different resolver locations spread out across The Netherlands. What is interesting is that the validation rate steadily rose until week 39 and then showed a significant drop.

I am – of course – very curious to find out what caused this drop. We use both the root trust anchor as well as ISC’s DLV repository to retrieve trust anchors that cannot be reached from the root down.

Anyway, I will try my best to send the first person who can give me a satisfactory explanation a nice pie ;-). Winner to be announced here.

2 Comments to “Puzzle of the week, win a pie :-)”

  1. Tweets die vermelden DNSSEC » Puzzle of the week, win a pie :-) -- Topsy.com says:

    […] Dit blogartikel was vermeld op Twitter door relaxnow, Roland van Rijswijk. Roland van Rijswijk heeft gezegd: Who can explain this drop in validation rate starting week 40? http://bit.ly/9O0GPk #dnssec #win-a-pie […]

  2. Hugo Salgado says:

    Assuming that you didn’t see failed validations in weeks 39 and 40, I guess that your statistics are counting twice when there’s a chain of validation in DLV and the root. So you have the case of .SE who added its DS to the root at the end of august, and then remove it from ITAR (and since DLV imports ITAR, from DLV) at the end of september. And I’m pretty sure that was the .GOV case too. I can see that if you take off the weeks 34-39, the lines are almost growing continously, don’t you think?
    So my bet is that you didn’t had a drop in week 40, but a double-validation counter in between ~34 and 39.