- ripe-atlas - mailman.ripe.net

Very short uptime for some probes
by Mark Delany 22 Nov '13

22 Nov '13

Hi all. I've only been on this list for a relatively short while so this is most likely in an FAQ I should be digging thru... I've noticed that the uptime for some of the physical probes we've installed are very short. According to the control panel, one in particular has connect times of around 2-3 minutes and disconnect times of 4-6 minutes. This cycle is repeated every 10 minutes or so. All day, every day. Superficially that sort of pattern suggests some sort of continuous reboot loop. However, if I ping the probe remotely, I get just an occasional packet loss. Certainly no long gaps that you'd expect from a continuous reboot. That I can ping it remotely also suggests that external connectivity isn't the problem. Besides the periodicity is just too regular and frequent for a routing loss or similar. (In any event, if the network it's on is losing connectivity every 5 minutes I would have plenty of other symptoms.) The problem probe is also reported as getting UDMs assigned and running them so I'm not quite sure how it manages to squeeze those tests in if that connectivity status report is accurate. So my main question is, should I be concerned about the control panel connectivity reports? And if I should be, is there much I can do beyond ensuring the device has good connectivity? Mark.

3 6

Categorize probes (feature request)
by Klaus Darilion 22 Nov '13

22 Nov '13

Hi! Depending on the measurement, sometimes I want a realistic "user's view" which means using probes in residential areas, and sometimes I want a "backbone" view, which mean using probes in the back-bone (good connected) e.g. to measure my access network or my servers without the delay introduced by the probes access network. Thus, it would be cool if the probe selection algorithm allows also to choose "residential" probes or "backbone" probes. Maybe this categorization can be done automatically, or probe owners can be asked to specify the probes connectivity manually. Thanks, Klaus

11 26

UI displays of measurements
by Gilles Massen 22 Nov '13

22 Nov '13

Hi, When you visualize a measurement in the GUI, what data is used for tabs like "Probes" or "Map"? Is that the latest available data from each participating probe, or something else? Gilles -- Fondation RESTENA - DNS-LU 6, rue Coudenhove-Kalergi L-1359 Luxembourg tel: (+352) 424409 fax: (+352) 422473

3 3

Change account email address
by Mike. 22 Nov '13

22 Nov '13

Hi, Is there a way to change the email address that is used to log in to the RIPE Atlas site? I am talking about the email address that is the answer to the prompt: Sign in using your RIPE NCC Access account and not the alert email address used when the probe goes offline for a period of time. Thanks, Mike.

4 5

Sudden raise in UDM credit consumption
by Tarko Tikan 20 Nov '13

20 Nov '13

hey, Has there been a change how UDM probe prices are calculated? I stopped old UDM (1004320) that had following parameters: - v4 ping - 10 probes - 240 seconds, 3 packets, size 500 - 3.0 credits per result, 10800 credits per day When I created identical UDM (1040516) (UDM parameters should really be changeable to there should be possiblity to clone UDM and then modify all parameters) with double amount of probes (20) and same parameters, I'm now suddenly charged 31.25 credits per result. This does not follow the linear logic as only parameter that changed is number of probes. -- tarko

2 2

Power plugs for the Anchor nodes
by Job Snijders 15 Nov '13

15 Nov '13

Dear all, A moment of clarity occurred to me, just before I shipped out some anchor nodes: the power plugs on the back of box (C7 type) won't easily be connected in most datacenter deployments. I found these cute plugs [1], which can be ordered here [2]. Hope this helps others :-) How have others solved the issue? Any ideas on how to get these Anchors to run on 48V DC (for telco-style deployment)? Kind regards, Job [1] - http://www.kempelektroniks.nl/FilesContent/menu/menu000250/images_iecc7adap… [2] - http://shop.kempelektroniks.nl/index.php/accessories/iec-adapters.html

4 3

probe allocations
by Gilles Massen 14 Nov '13

14 Nov '13

Hi, I was wondering how the number of allocated probes is determined for a given UDM. Specifically, I set up a recurring UDM with 50 world-wide probes (DNS), but 110 probes were allocated to the test. While I'm certainly flattered that RIPE Atlas deemed the measurement so worthwhile, the credit consumption was somewhat higher than expected... so what is the logic, and am I able to get (at most) what I requested? best, Gilles -- Fondation RESTENA - DNS-LU 6, rue Coudenhove-Kalergi L-1359 Luxembourg tel: (+352) 424409 fax: (+352) 422473

3 10

Bulk access to the measurements?
by Stephane Bortzmeyer 13 Nov '13

13 Nov '13

A colleague asks me if it is possible to have bulk access to public measurement data. (Properly: otherwise, he could dor a for loop over all measurement IDs but I suspect it is not nice.) I find nothing in the Atlas documentation. Is it because I do not search hard enough or because this possibility does not exist? #BigData

4 3

Archiving user-defined measurements in Atlas
by ivo.vdmaagdenberg＠gmail.com 12 Nov '13

12 Nov '13

Hi, Played a bit with measurements, now I want make some sensible ones and get rid of the rest. In the documentation "How do I stop a user-defined measurement?" there is mention of the 'Archive' option per measurement. I see that option but it is grayed out. The question is: What are the circumstance that grey the option? This holds for the "Forced to Stop", "Stopped" and "Ongoing" UDMs (maybe more...) Kind regards, Ivo

2 1

Feature request: Regularly Publish Names
by Daniel Karrenberg 12 Nov '13

12 Nov '13

Title: Regularly Publish Names Description: Regularly collect all IP addresses referenced in measurement definitions and all results. Reverse resolve the addresses in the DNS. Collect netnames for the addresses from the RIR registration data. Possibly collect other data such as geo-location and ASes announcing. Publish this collected data both in bulk and thru the API indexed by IP. A frequency of once every 24 hours would be sufficient to stat with. Use: When analysing measurements from the past it is important to have meta information that was current at the time of the measurement. Having this data collected in advance often also allows for more efficient analysis and more responsive tools. Perceived complexity: Low. Use hadoop to collect the list of IPs. Then run a low priority job to collect the meta data. The job also has high potential for parallelism. Perceived cost: Low. Just some cycles and storage on the back end.

1 0