- ripe-atlas - mailman.ripe.net

Software probe keeps going to disconnected
by vom513 19 Jun '24

19 Jun '24

Hello all, I have 2 software probes. One is on a datacenter VPS and the other at home (runs on Proxmox). The one at home keeps going to disconnected. The probe is Debian 12. Probe version is 5080-1. This machine has v6 as well - so I was thinking maybe something with temp/privacy addresses getting reaped but no, I only have one good old EUI-64 address on there. (Knock on wood) - no other VMs on my Proxmox machine have any issues like this. The only thing that looks remotely related (and this happened a bit before Nagios sent out an alert) is this: Jun 17 12:34:41 atlas-probe ATLAS[123749]: /usr/local/atlas/bin/ATLAS: 99: kill: No such process Jun 17 12:34:41 atlas-probe ATLAS[123749]: no ssh client matching . cleanup state files. for next restart Jun 17 12:34:41 atlas-probe ATLAS[123749]: RESULT 9006 done 1718642081 bc2411ca94c6 no reginit.vol start registration Jun 17 12:34:41 atlas-probe ATLAS[123749]: /var/atlas-probe/status/reginit.vol does not exist try new reg Jun 17 12:34:45 atlas-probe ATLAS[123749]: Ping works Jun 17 12:34:45 atlas-probe ATLAS[123749]: /usr/local/atlas/bin/ATLAS: 92: kill: No such process Jun 17 12:34:45 atlas-probe ATLAS[123749]: start reg Jun 17 12:34:45 atlas-probe ATLAS[144349]: ATLAS registration starting Jun 17 12:34:45 atlas-probe ATLAS[144349]: REREG_TIMER 1718416507 expired now is 1718642085 Jun 17 12:34:45 atlas-probe ATLAS[144349]: REREG_TIMER_EXPIRED go re register REREG_TIMER 1 , now is 1718642085 Jun 17 12:34:45 atlas-probe ATLAS[144349]: REGHOSTS reg03.atlas.ripe.net 193.0.19.246 2001:67c:2e8:11::c100:13f6 reg04.atlas.ripe.net 193.0.19.247 2001:67c:2e8:11::c100:13f7 Jun 17 12:34:45 atlas-probe ATLAS[144349]: ssh -p 443 atlas(a)reg03.atlas.ripe.net INIT Jun 17 12:34:48 atlas-probe ATLAS[144349]: Got good controller info Jun 17 12:34:48 atlas-probe ATLAS[144349]: check cached controller info from previous registration Jun 17 12:34:48 atlas-probe ATLAS[144349]: NO cached controller info. NO REMOTE port info Jun 17 12:34:48 atlas-probe ATLAS[144349]: Do a controller INIT Jun 17 12:34:48 atlas-probe ATLAS[144349]: Controller init -p 443 atlas(a)ctr01.atlas-prod.aws.ripe.net INIT I believe I should see an persistent ssh outbound ? The only ESTAB is me ssh’ing in, and a 443 to RIPE: root@atlas-probe:/home/vom# ss -planto | grep ESTAB ESTAB 0 0 192.168.69.111:22 192.168.64.21:65409 users:(("sshd",pid=145083,fd=4),("sshd",pid=145059,fd=4)) timer:(keepalive,115min,0) ESTAB 0 0 [2603:6011:6300:1f02:be24:11ff:feca:94c6]:37176 [2a13:27c0:1c:d01:cc61:f3a3:920d:cd88]:443 users:(("ssh",pid=144380,fd=3)) timer:(keepalive,103min,0) Let me know what else I can look at. I assume there is some debug I can crank up ? Just going to reboot this for now to get it happy and get Nagios to shut up. Thanks.

2 2

[ anchor] Quota on the usage of an anchor in measurement.
by Huseyn Gambarov 19 Jun '24

19 Jun '24

Hello. In my research, I am conducting multiple RTT measurements against different targets. During the measurement process, a single anchor stops getting scheduled for a requested measurement after some number of times or returns -1 as a result. I want to ask if there is a limit on how many times a single anchor can participate in a measurement. And what is its cooldown period if such a limit exists? With regards, Huseyn Gambarov Research Assistant PhD Student in Computer Science Case Western Reserve University

2 1

Atlas UI updated
by Stephen Suess 10 Jun '24

10 Jun '24

Hi Atlas Users, Along with the aforementioned removal of old measurement detail page links, we have just released an update to the UI that includes: - NEW: Pref for using local time (vs UTC) everywhere - NEW: Added local toggle to QuickLook for using probes close to user location - CHANGE: Add traceroute table to msm overview map marker popups - CHANGE: Make clearer the difference between no report (yet) vs no reply (from target) - CHANGE: Map now resizes on larger screens in measurement overview If you want to see the full list of what has changed, you can check the release notes here: https://atlas.ripe.net/docs/releases/website/2024-06-10.html As always, we hope you find these changes useful and we welcome your feedback. Kind regards, Stephen Atlas UI

1 0

Removing links to legacy measurement detail pages
by Stephen Suess 07 Jun '24

07 Jun '24

Dear Atlas Users, Now that we have been up and running with the new measurement detail pages for a few weeks, we plan to remove the banner and links to the legacy pages next week. Please let us know if you still have any need for the old detail pages and why, and we will do our best to address any concerns. Kind regards, Stephen Atlas UI

1 0

RIPE Atlas infrastructure maintenance on 11 June 2024
by Robert Kisteleki 07 Jun '24

07 Jun '24

Dear RIPE Atlas users, On 11 June 2024, between 8:30-11:30 (CEST) we will carry out maintenance of the central RIPE Atlas infrastructure. The UI, APIs, access to stored results and metadata, as well as measurement scheduling will be unavailable. During this time results will still be still collected and stored. We apologise for the inconvenience this may cause. Regards, Robert Kisteleki, for the RIPE Atlas team

1 0

How to be more helpful...
by Mike 01 Jun '24

01 Jun '24

OK, here's my "problem." For starters I think the Atlas project is a great thing for the Internet, and I thank my local ISP, Comcast (USA), for encouraging me to sign up. But here's the thing for me. I have so many millions of points accumulated. When I see a message on this message list about someone asking for, say 20 million points, by the time I see that message, the requester has already received, typically, more than twice the number of points requested. (and I will say right now, private requests for points will be ignored) So my question becomes... I participate in this amazing project with a measurement device. But I also want to participate in the project by helping others, I guess I am asking, is there a way I can contribute points without having to check the list every hour or so?

2 1

Atlas UI updated
by Stephen Suess 28 May '24

28 May '24

Hi Atlas fans, We got a lot of useful feedback at RIPE88, and have just released some changes to production. Here are some of the highlights: - NEW: Pref for setting Mine tab as default for lists when logged in (instead of Public) - CHANGE: Improved sorting of minRTTs, no reply, n/a - CHANGE: Measurement form now using modal popups for probe selection to clarify interaction - CHANGE: When adding probes in the measurement form, it will now ask you if you want to keep the default probes or remove them - CHANGE: Rework map legend enable/disable logic when clicking on labels - CHANGE: RD bit set true by default for new DNS measurements If you want to see the full list of what has changed, you can check the release notes here: https://atlas.ripe.net/docs/releases/website/2024-05-28.html As always, we hope you find these changes useful and we welcome your feedback. Kind regards, Stephen Atlas UI

1 0

TraceMON pages broken?
by Malte Tashiro 28 May '24

28 May '24

Hi, it seems like the TraceMON widget is currently broken. The tracemon-dist.js [0] file appears to be missing, which is required by the main traceroute-widget-main.js [1] file. Tried with Chromium and Firefox and both the old [2] and new [3] way. Would be happy about a fix :) Best, Malte [0] https://www-static.ripe.net/static/rnd-ui/atlas/static/measurements/widgets… [1] https://www-static.ripe.net/static/rnd-ui/atlas/static/measurements/widgets… [2] https://atlas.ripe.net/measurements/9333344/#tracemon [3] https://atlas.ripe.net/tracemon/9333344

2 1

Delays?
by Ernst J. Oud 14 May '24

14 May '24

I experience delays on creating measurements. Even after 2 minutes, a measurement is not running, eventually it will but it used to take a couple of seconds. My streaming api calls now time out… Is this known? Will it approve or do I need to change my scripts so that they wait longer? Regards, Ernst J. Oud

3 2

API advice to pull data from multiple measurements
by techtebatoye＠gmail.com 14 May '24

14 May '24

Hi, I would like to pull data from DNSMON probe measurements to build a real-time view. Considering the combination of IPv4/IPv6, UDP/TCP, multiple nameservers per TLD, it quickly adds up to 150+ measurements. I started with the streaming API, sending register messages for the measurements of interest, but quite quickly I am getting errors with "slow_down" response. I could not find the rate limit values in the documentation of the streaming API. Using the HTTP API would also mean hundreds of API calls, one per measurement. What is the recommended way to regularly pull data from many measurements, to avoid rate limiting or generating too much load on the API ? Regards

2 1