Skip to main content

You're viewing an archived page. It is no longer being updated.

RIPE Database Working Group Minutes RIPE 80

Thursday, 14 May 15:00 - 15:45
Chairs: William Sylvester, Denis Walker
Scribe: Oleg Muravskiy
Status: Draft

A. Introduction   

William welcomed attendees.

B. Operational Update RIPE Database

Edward Shryane, RIPE NCC

This presentation is available at:

Cynthia Revström thanked the RIPE NCC’s Database Team for fixing the “huge update” bug and confirmed that she had verified in the testing instance that it did appear to be fixed.

C. Cloud Migration

Edward Shryane, RIPE NCC   

This presentation is available at:

Robert Scheck, ETES, asked about the current setup – how many physical systems were they talking about and whether virtualisation was already being used.

Ed said they were using virtualisation extensively for their test and release candidate environments (and had been for a number of years), so the team had had a lot of experience with this. Regarding their physical production infrastructure, he said he could share some details about this on the mailing list.

John Bond, Wikimedia Foundation, asked if there was a shortlist of the cloud providers being considered.

Ed said the decision hadn’t been made yet, but the team would probably be looking at Amazon and Google Cloud, which were already being used by other departments within the RIPE NCC.

Cynthia Revström asked if they had considered open-sourcing their Webupdates interface.

Ed said this was definitely something they were considering. This could be useful, as the community could reuse some of the functionality and for developers it was preferable to work with open code. It would also provide possibilities for audit, as users could more easily see how things work and report bugs.

Daniel Piekacz, Twitch, asked if there was a plan to maintain internal scaled down infrastructure for the unlikely event of a cloud provider outage, and if this would be a hybrid cloud solution.

Ed said this was indeed the plan. Normally traffic would be served from the cloud, but for disaster recovery they would fall back to virtualised infrastructure internally.

D. PII Privacy Leak in descr: attributes

Michael Kafka, NetCaT GmbH

The presentation is available at:

Maria Stafyla, RIPE NCC Senior Legal Counsel, said the "descr:" attribute was not intended to contain personal data and whoever inserted personal data into the RIPE Database was responsible for ensuring that the relevant legal conditions were met. However, the RIPE NCC had a process that allowed for the removal of personal data from the RIPE Database (described at:

Michael replied that the end user had direct contact with the LIR, not the RIPE NCC, and the removal process was quite complicated. He recommended they implement a mechanism that would allow the LIR to remove historical personal information.

Maria added that it was also possible to submit a request for removal by sending an email to [email protected] with the relevant details. Personal data in the "descr:" attribute was not filtered because it was never meant to contain personal data in the first place, though it could often contain information that was useful for historical or research purposes. The intention there was to find a balance between satisfying the community's requirement for historical registry information and privacy restrictions.

George Michaelson, APNIC, thanked Michael for his presentation, which in his opinion described a problem for any NRTM feed receiver that maintained history state.

E. NWI - Open Proposals

Denis Walker, co-chair

The presentation is available at:

George commented that RRDP had huge advantage that was well tuned to CDN distribution.

William suggested they move this discussion to the mailing list.

Z. AOB (open discussion)

There were no AOBs.

End of session.