[dns-wg] Incorrect RCODE values in DNSMON and DomainMon

Dear colleagues,

In May I emailed the list (see below) about incorrect RCODE values in 
DNSMON and DomainMon. Since then we fixed the bug that was causing this 
issue.

The fix has been in place since the beginning of July. However, 
correcting the RCODE values in older data is taking more time than 
expected. We now expect all of the (historic) RCODE values to be 
corrected by the end of August.

As mentioned in the previous announcement, there is no impact on the 
actual measurement data itself.

I apologise for any inconvenience caused.

Kind regards,
Romeo Zwart




----- original message -------


Dear colleagues,

It has been brought to our attention that the RCODE values resulting
from DNS measurements in RIPE Atlas are sometimes presented incorrectly
in DNSMON and DomainMon. We are currently examining this in more detail,
but preliminary investigations suggest that this is due to a bug in the
software parsing and storing DNS measurement results in our backend systems.

Note that there is no impact on the DNS measurement data itself - only
the RCODE values are affected.

This issue has a potential impact on DNSMON and DomainMon visualisations
if one filters on RCODE by checking the "Exclude errors" tick box. This
filter is not enabled by default, so it will not normally have an impact
on these visualisations.

We are still investigating the full extent of this problem. Based on the
information available, we expect to have fixed the issue and corrected
the historic RCODE values by the end of June.

We will update you via this list when we have more information and a
more accurate ETA for the fix.

Kind regards,

Romeo Zwart