Malte,

Yes, I noticed the same thing. I have a long running measurement for which I regularly check to include new or exclude disconnected probes.

As a result it now shows 177 requested and 29 participating, but it currently shows 34 measurement results (https://atlas.ripe.net/measurements/45518789/#general).

It does not at all add up.

Regards,

Ernst J. Oud


Met vriendelijke groet,

Ernst J. Oud
On 8 Apr 2024, at 08:30, Malte Tashiro via ripe-atlas <ripe-atlas@ripe.net> wrote:

Hi,

I'm currently looking at measurement metadata and am a bit confused about the different ways of getting the number of probes participating in a measurement and the number of probes requested by the user.

I'd like to know how to...
 1. get the number of probes _currently_ participating in a measurement (for ongoing measurements)
 2. the number of probes that participated when the measurement finished (for stopped measurements)
 3. the number of probes a user requested over the lifetime of a measurement

The default response to /measurements returns a participant_count and a probes_requested value. However, there are also the optional fields current_probes and participation_requests and these do not add up.

For example, this [0][1] measurement (not mine) has a participant_count of 0 (showing as "Actually Participating: None") in the web interface, but one probe listed in current_probes, which shows up in the "Latest Results" list and is still performing measurements (at least while writing this).
So it seems like the current_probes list is more accurate to get the number of probes _currently_ (or when the measurement finished) participating in a measurement?

Similarly, there are 35 probes_requested, which equals the sum of "requested" probes from the participation_requests, however, some request are actually "remove" actions, which should not be included in this count?

Any hints on how to interpret these fields?

I think the requested probes and participation requests are more a problem for long-running measurements, but the participant count is sometimes weird for one-offs as well.

Thanks,
Malte

[0] https://atlas.ripe.net/measurements/67903255
[1] https://atlas.ripe.net/api/v2/measurements/67903255?optional_fields=current_probes,participation_requests
<OpenPGP_0x7D82498BEF2E08F8.asc>
--
ripe-atlas mailing list
ripe-atlas@ripe.net
https://lists.ripe.net/mailman/listinfo/ripe-atlas