From lesnix at gmail.com Tue Jan 4 10:38:38 2011 From: lesnix at gmail.com (Egor Zimin) Date: Tue, 04 Jan 2011 12:38:38 +0300 Subject: [atlas]Probe actually down? In-Reply-To: <936CFCAC-F864-4FD4-AA99-2619D51A04C5@bbn.com> References: <936CFCAC-F864-4FD4-AA99-2619D51A04C5@bbn.com> Message-ID: <4D22EA9E.7040907@gmail.com> I see the same situation with probe #163. 30.12.2010 22:00, Richard L. Barnes ?????: > Probe #394 is reported as being down in the web GUI, but I can ping6 to it without a problem. I'm not near the probe right now, so I can't look at it or its traffic. Are there known cases where probes are alive but not reporting to their controllers? > > --Richard -- Best regards, Egor Zimin From Piotr.Strzyzewski at polsl.pl Tue Jan 4 10:46:02 2011 From: Piotr.Strzyzewski at polsl.pl (Piotr Strzyzewski) Date: Tue, 4 Jan 2011 10:46:02 +0100 Subject: [atlas]Probe actually down? In-Reply-To: <4D22EA9E.7040907@gmail.com> References: <936CFCAC-F864-4FD4-AA99-2619D51A04C5@bbn.com> <4D22EA9E.7040907@gmail.com> Message-ID: <20110104094602.GA7858@hydra.ck.polsl.pl> On Tue, Jan 04, 2011 at 12:38:38PM +0300, Egor Zimin wrote: > I see the same situation with probe #163. Maybe this is the right time to implement optional mail notifications about such events? Piotr -- gucio -> Piotr Strzy?ewski E-mail: Piotr.Strzyzewski at polsl.pl From robert at ripe.net Wed Jan 5 11:22:39 2011 From: robert at ripe.net (Robert Kisteleki) Date: Wed, 05 Jan 2011 11:22:39 +0100 Subject: [atlas]Probe actually down? In-Reply-To: <936CFCAC-F864-4FD4-AA99-2619D51A04C5@bbn.com> References: <936CFCAC-F864-4FD4-AA99-2619D51A04C5@bbn.com> Message-ID: <4D24466F.9080804@ripe.net> Replying to various mails in the thread: On 2010.12.30. 20:00, Richard L. Barnes wrote: > Probe #394 is reported as being down in the web GUI, but I can ping6 to it without a problem. I'm not near the probe right now, so I can't look at it or its traffic. Are there known cases where probes are alive but not reporting to their controllers? > > --Richard There are a few scenarios when this is possible: * the probe is powered on, but it's not connected to the infrastructure (for example the connection went down for some reason, or after a reboot). In this case you can still ping it as the network stack is running, but it's reported as "down". * Every now and then we have issues with the controllers' network connections. They do reconnect themselves reliably, but if a probe connects to such a controller in the meantime, then the controller cannot report the probe status change to our database. A fix is in the making for this. On 2011.01.04. 10:38, Egor Zimin wrote: > I see the same situation with probe #163. This probe seems to be up now. There are days when it fluctuates, but also periods when it's up for days. On 2011.01.04. 10:46, Piotr Strzyzewski wrote: > Maybe this is the right time to implement optional mail notifications > about such events? So you'd like to subscribe to notifications when your probe comes up or goes down? It's certainly possible to do this, but -- since probe reconnects are not really that serious events -- I think I'd rather want be notified if a probe was down for a non-trivial time. Cheers, Robert From robert at ripe.net Wed Jan 5 11:29:37 2011 From: robert at ripe.net (Robert Kisteleki) Date: Wed, 05 Jan 2011 11:29:37 +0100 Subject: [atlas]trying to collect basic info about the architecture In-Reply-To: <4D11DC68.8070906@ripe.net> References: <4D10E28D.5060708@CC.UniVie.ac.at> <4D11DC68.8070906@ripe.net> Message-ID: <4D244811.7000507@ripe.net> Some additional info: On 2010.12.22. 12:09, Emile Aben wrote: > On 21/12/2010 18:23, Wilfried Woeber, UniVie/ACOnet wrote: >> Is the probe capable of bufffering data for a while if connectivity to the >> Net is still there, but not to the controller(s)? > > Yes, the probes have a limited capability to do so. I think it's > currently capable of storing a couple of minutes worth of data. It could store more, but we're cautious for now as in the worst case this could fill up the storage space which could cause serious problems. We'll add more buffering as soon as we're confident it cannot fire back. >> Is there any means to (locally, on the same subnet, or even remotely) >> trigger or "reboot" a probe, other than disconnecting power (which means >> physical access required)? > > Currently there is no way for users to reboot a probe other then > disconnecting power. If this is a capability that people would like to > have, we could look into ways we could make that happen. This makes me wonder: what problem would this solve? You obviously couldn't use this feature if the probe was down, only if it was up; but then why would you want to reboot it? :-) Cheers, Robert From Piotr.Strzyzewski at polsl.pl Wed Jan 5 11:31:49 2011 From: Piotr.Strzyzewski at polsl.pl (Piotr Strzyzewski) Date: Wed, 5 Jan 2011 11:31:49 +0100 Subject: [atlas]Probe actually down? In-Reply-To: <4D24466F.9080804@ripe.net> References: <936CFCAC-F864-4FD4-AA99-2619D51A04C5@bbn.com> <4D24466F.9080804@ripe.net> Message-ID: <20110105103149.GC25130@hydra.ck.polsl.pl> On Wed, Jan 05, 2011 at 11:22:39AM +0100, Robert Kisteleki wrote: > On 2011.01.04. 10:46, Piotr Strzyzewski wrote: > > Maybe this is the right time to implement optional mail notifications > > about such events? > > So you'd like to subscribe to notifications when your probe comes up or goes > down? It's certainly possible to do this, but -- since probe reconnects are > not really that serious events -- I think I'd rather want be notified if a > probe was down for a non-trivial time. This sugestion looks reasonable. Piotr -- gucio -> Piotr Strzy?ewski E-mail: Piotr.Strzyzewski at polsl.pl From Tjebbe at kanariepiet.com Wed Jan 5 11:43:31 2011 From: Tjebbe at kanariepiet.com (Jelte) Date: Wed, 05 Jan 2011 11:43:31 +0100 Subject: [atlas]trying to collect basic info about the architecture In-Reply-To: <4D244811.7000507@ripe.net> References: <4D10E28D.5060708@CC.UniVie.ac.at> <4D11DC68.8070906@ripe.net> <4D244811.7000507@ripe.net> Message-ID: <4D244B53.30800@kanariepiet.com> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 01/05/2011 11:29 AM, Robert Kisteleki wrote: >> Currently there is no way for users to reboot a probe other then >> disconnecting power. If this is a capability that people would like to >> have, we could look into ways we could make that happen. > > This makes me wonder: what problem would this solve? You obviously couldn't > use this feature if the probe was down, only if it was up; but then why > would you want to reboot it? :-) > Well, there have been several cases of probes being up but not reporting any data. At least one of those cases was solved by a reboot :) Of course fixing all the issues that causes such behaviour is the ultimate goal, but since every piece of code contains at least one bug, i can see some use for such a feature for those people having their probes not in an easily reachable location. Having some form of remote control over one's usb power would also do the trick ;) Jelte -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAk0kS1MACgkQ4nZCKsdOncXT1gCaA/ILH87xVGDHKF9n0wKXgDsx DLEAoJw1ksFf8viujiGLhVnV+7SWIFKV =w3/Q -----END PGP SIGNATURE----- From robert at ripe.net Wed Jan 5 12:12:21 2011 From: robert at ripe.net (Robert Kisteleki) Date: Wed, 05 Jan 2011 12:12:21 +0100 Subject: [atlas]trying to collect basic info about the architecture In-Reply-To: <4D244B53.30800@kanariepiet.com> References: <4D10E28D.5060708@CC.UniVie.ac.at> <4D11DC68.8070906@ripe.net> <4D244811.7000507@ripe.net> <4D244B53.30800@kanariepiet.com> Message-ID: <4D245215.1070109@ripe.net> On 2011.01.05. 11:43, Jelte wrote: > On 01/05/2011 11:29 AM, Robert Kisteleki wrote: >>> Currently there is no way for users to reboot a probe other then >>> disconnecting power. If this is a capability that people would like to >>> have, we could look into ways we could make that happen. > >> This makes me wonder: what problem would this solve? You obviously couldn't >> use this feature if the probe was down, only if it was up; but then why >> would you want to reboot it? :-) > > > Well, there have been several cases of probes being up but not reporting any > data. At least one of those cases was solved by a reboot :) Good point, but that's not the expected behavior :-) > Of course fixing all the issues that causes such behaviour is the ultimate goal, > but since every piece of code contains at least one bug, i can see some use for > such a feature for those people having their probes not in an easily reachable > location. I agree, but my argument is that in normal circumstances this isn't useful. Also, we can always force probes to reconnect (by kicking them out from the server side). Besides, if we implemented this feature, we'd have less incentives to solve the bugs :-) Cheers, Robert > Having some form of remote control over one's usb power would also do the trick ;) > > Jelte From robert at ripe.net Wed Jan 12 10:22:32 2011 From: robert at ripe.net (Robert Kisteleki) Date: Wed, 12 Jan 2011 10:22:32 +0100 Subject: [atlas]Some new UI features Message-ID: <4D2D72D8.9040901@ripe.net> Hello, We've added some changes to the UI yesterday. From now on upon registering a new probe, or changing its attributes (from the probe status page) the host can set the following additional information: * Router type: optional, describes the router/NAT model you have, mainly in a home installation. This information could be useful for us to correlate weird probe behavior across similar router models. * Allowed bandwidth: it's not used yet, but you can imagine the future purpose for it :) * DNS options: this has been asked by some of you before. By default it's off. If you set it to "simple" then your probe will have a probe-xxxx.probes.atlas.ripe.net DNS entry. If you set it to "obfuscated" then your probe will have a .probes.atlas.ripe.net DNS entry. We'll activate this feature soon. * Public: you can make your probe visible to everyone, which means that all Atlas users will see the probe in their probe list (with role "Viewer"), and they will be able to see the measurement details on the probe status page. Consider this an experimental feature -- we have to see if it's useful or not. You'll need to reload the page (with the browser "reload") in order to see these changes. Regards, Robert From Woeber at CC.UniVie.ac.at Wed Jan 12 14:42:13 2011 From: Woeber at CC.UniVie.ac.at (Wilfried Woeber, UniVie/ACOnet) Date: Wed, 12 Jan 2011 13:42:13 +0000 Subject: [atlas]Some new UI features In-Reply-To: <4D2D72D8.9040901@ripe.net> References: <4D2D72D8.9040901@ripe.net> Message-ID: <4D2DAFB5.3070908@CC.UniVie.ac.at> Robert Kisteleki wrote: > Hello, > > We've added some changes to the UI yesterday. From now on upon registering a > new probe, or changing its attributes (from the probe status page) the host > can set the following additional information: > > * Router type: optional, describes the router/NAT model you have, mainly in > a home installation. This information could be useful for us to correlate > weird probe behavior across similar router models. Great idea! > * Allowed bandwidth: it's not used yet, but you can imagine the future > purpose for it :) :-) > * DNS options: this has been asked by some of you before. By default it's > off. If you set it to "simple" then your probe will have a > probe-xxxx.probes.atlas.ripe.net DNS entry. If you set it to "obfuscated" > then your probe will have a .probes.atlas.ripe.net > DNS entry. We'll activate this feature soon. > > * Public: you can make your probe visible to everyone, which means that all > Atlas users will see the probe in their probe list (with role "Viewer"), and > they will be able to see the measurement details on the probe status page. > Consider this an experimental feature -- we have to see if it's useful or not. Well, I'll take a short break before I set the tick-mark for public :-) I think it is a nice idea, too, with 2 minor caveats: - the probes with type host should remain at the top of the list (the sort function for the column seem to not being available yet), and - rather than "public", or as an alternative, I would prefer to enable known users to beceome "viewers" and even be allowed to change configuration. This would be very helpful for en environment like ours, where there's more than one host for our probes. Right now both these numbers are *very* small :-), but if this develops into the structure I'd like to see, it will no longer be easy... > You'll need to reload the page (with the browser "reload") in order to see > these changes. > > Regards, > Robert Thanks a lot, Wilfried. From taihen at taihen.org Wed Jan 12 15:12:24 2011 From: taihen at taihen.org (Andrzej Wolski) Date: Wed, 12 Jan 2011 15:12:24 +0100 Subject: [atlas]Not my probes ? Message-ID: Hi everyone, after coming back from holidays I've looked at the graphs of my probe ( 00:20:4A:C8:24:17 ) and found that I have rights to view about 16 other probes from all over the place. Bug or a feature ? ;-) -- Andrzej Wolski aka Taihen mailto/xmpp: taihen at taihen.org From robert at ripe.net Wed Jan 12 15:51:02 2011 From: robert at ripe.net (Robert Kisteleki) Date: Wed, 12 Jan 2011 15:51:02 +0100 Subject: [atlas]Not my probes ? In-Reply-To: References: Message-ID: <4D2DBFD6.8000902@ripe.net> On 2011.01.12. 15:12, Andrzej Wolski wrote: > Hi everyone, > > after coming back from holidays I've looked at the graphs of my probe > ( 00:20:4A:C8:24:17 ) and found that I have rights to view about 16 > other probes from all over the place. > > Bug or a feature ? ;-) It is intentional, and we like to think about it as a feature :-) I sent a mail to the list today about it - and already more than a dozen probes have been made available! Cheers, Robert From rbarnes at bbn.com Wed Jan 12 16:25:49 2011 From: rbarnes at bbn.com (Richard L. Barnes) Date: Wed, 12 Jan 2011 10:25:49 -0500 Subject: [atlas]Some new UI features In-Reply-To: <4D2D72D8.9040901@ripe.net> References: <4D2D72D8.9040901@ripe.net> Message-ID: > * Public: you can make your probe visible to everyone, which means that all > Atlas users will see the probe in their probe list (with role "Viewer"), and > they will be able to see the measurement details on the probe status page. > Consider this an experimental feature -- we have to see if it's useful or not. This is a nice option, but putting all the public probes in the "My Probes" window makes things a little cluttered. Perhaps you could make a separate "Public Probes" tab? --Richard From joao at bondis.org Wed Jan 12 16:26:54 2011 From: joao at bondis.org (=?iso-8859-1?Q?Jo=E3o_Damas?=) Date: Wed, 12 Jan 2011 16:26:54 +0100 Subject: [atlas]Some new UI features In-Reply-To: References: <4D2D72D8.9040901@ripe.net> Message-ID: <1FB937C4-6CAE-4240-9630-25374C8356B7@bondis.org> Strong +1, couldn't spot my probes amongst all other my-probes Joao On 12 Jan 2011, at 16:25, Richard L. Barnes wrote: >> * Public: you can make your probe visible to everyone, which means that all >> Atlas users will see the probe in their probe list (with role "Viewer"), and >> they will be able to see the measurement details on the probe status page. >> Consider this an experimental feature -- we have to see if it's useful or not. > > This is a nice option, but putting all the public probes in the "My Probes" window makes things a little cluttered. Perhaps you could make a separate "Public Probes" tab? > > --Richard From robert at ripe.net Wed Jan 12 16:34:43 2011 From: robert at ripe.net (Robert Kisteleki) Date: Wed, 12 Jan 2011 16:34:43 +0100 Subject: [atlas]Some new UI features In-Reply-To: <1FB937C4-6CAE-4240-9630-25374C8356B7@bondis.org> References: <4D2D72D8.9040901@ripe.net> <1FB937C4-6CAE-4240-9630-25374C8356B7@bondis.org> Message-ID: <4D2DCA13.4060604@ripe.net> OK, we'll think about something, but in the meantime the 'Role' column is your friend -- it's easy to spot the outlier there, and you can also filter to "Host". Robert On 2011.01.12. 16:26, Jo?o Damas wrote: > Strong +1, couldn't spot my probes amongst all other my-probes > > Joao > > On 12 Jan 2011, at 16:25, Richard L. Barnes wrote: > >>> * Public: you can make your probe visible to everyone, which means that all >>> Atlas users will see the probe in their probe list (with role "Viewer"), and >>> they will be able to see the measurement details on the probe status page. >>> Consider this an experimental feature -- we have to see if it's useful or not. >> >> This is a nice option, but putting all the public probes in the "My Probes" window makes things a little cluttered. Perhaps you could make a separate "Public Probes" tab? >> >> --Richard > From rick at bytel.net.uk Wed Jan 12 16:32:23 2011 From: rick at bytel.net.uk (Rick Hodger) Date: Wed, 12 Jan 2011 15:32:23 -0000 Subject: [atlas]Some new UI features In-Reply-To: <1FB937C4-6CAE-4240-9630-25374C8356B7@bondis.org> References: <4D2D72D8.9040901@ripe.net> <1FB937C4-6CAE-4240-9630-25374C8356B7@bondis.org> Message-ID: <015e01cbb26d$e88bb340$b9a319c0$@net.uk> You can click on the Roles column and set the filter to 'host'. That removes all the public probes. It seems to remember it between login sessions too. Rick Hodger Senior ICT Administrator Bytel T/AS Atlas-Communications Ltd. > -----Original Message----- > From: ripe-atlas-admin at ripe.net [mailto:ripe-atlas-admin at ripe.net] On Behalf > Of Jo?o Damas > Sent: 12 January 2011 15:27 > To: Richard L. Barnes > Cc: Robert Kisteleki; ripe-atlas at ripe.net > Subject: Re: [atlas]Some new UI features > > Strong +1, couldn't spot my probes amongst all other my-probes > > Joao > > On 12 Jan 2011, at 16:25, Richard L. Barnes wrote: > > >> * Public: you can make your probe visible to everyone, which means that all > >> Atlas users will see the probe in their probe list (with role "Viewer"), > and > >> they will be able to see the measurement details on the probe status page. > >> Consider this an experimental feature -- we have to see if it's useful or > not. > > > > This is a nice option, but putting all the public probes in the "My Probes" > window makes things a little cluttered. Perhaps you could make a separate > "Public Probes" tab? > > > > --Richard From Woeber at CC.UniVie.ac.at Wed Jan 12 23:41:01 2011 From: Woeber at CC.UniVie.ac.at (Wilfried Woeber, UniVie/ACOnet) Date: Wed, 12 Jan 2011 22:41:01 +0000 Subject: [atlas]Some new UI features In-Reply-To: <4D2D72D8.9040901@ripe.net> References: <4D2D72D8.9040901@ripe.net> Message-ID: <4D2E2DFD.2060603@CC.UniVie.ac.at> Robert Kisteleki wrote: Another entry for the wish-list: :-) > * Public: you can make your probe visible to everyone, which means that all > Atlas users will see the probe in their probe list (with role "Viewer"), and it would be cool to have those probes tagged, which are IPv6 rnabled! Thanks, Wilfried. PS: and please add a segment for labs.ripe.net over IPv6! From mbehring at cisco.com Thu Jan 13 09:18:55 2011 From: mbehring at cisco.com (Michael Behringer (mbehring)) Date: Thu, 13 Jan 2011 09:18:55 +0100 Subject: [atlas]Probe comparisons Message-ID: <51865E1A72F61D48A247348DCCBA5CFB02E6A69B@XMB-AMS-105.cisco.com> Team, The new feature of making probes generally visible is EXCELLENT. It allows nice comparisons between different probes. For example, I just compared my Orange DSL probe to a Free probe (both France). The results are VERY interesting! (Would be interested to chat with the owner of this Free probe off line, BTW - can you ping me please?). Now of course this leads immediately to a new feature request! :-) (the curse of being popular!) It would be nice to be able to compare a number of probes side by side. In the simplest case, min/avg/max values (graphs are nice but not essential). Now of course the next feature request :-) that'll come out of this is to see the full traceroute from each probe, to be able to understand where the difference in RTT is coming from. On the positive side, the beer supply to the developers at the next RIPE meeting should be plenty! :-) Thanks for all your work! Michael From florian at narrans.de Thu Jan 13 10:18:24 2011 From: florian at narrans.de (Florian Obser) Date: Thu, 13 Jan 2011 10:18:24 +0100 Subject: [atlas]Some new UI features In-Reply-To: <4D2D72D8.9040901@ripe.net> References: <4D2D72D8.9040901@ripe.net> Message-ID: <4D2EC360.7060903@narrans.de> Hi, On 01/12/2011 10:22 AM, Robert Kisteleki wrote: > Hello, [...] > * Public: you can make your probe visible to everyone, which means that all > Atlas users will see the probe in their probe list (with role "Viewer"), and > they will be able to see the measurement details on the probe status page. > Consider this an experimental feature -- we have to see if it's useful or not. I see that some people put uplink details in the description field (as did I...). However since this is a text field the information is rather chaotic. I think we need a select box for this information, dsl, cable, 10 GE fibre to the home, neutrino transmission... No, I don't have an exhaustive list to put in the select box, if something like that isn't readily available maybe we can do a poll on this list. Wilfried suggested to have a tag for IPv6 enabled probes, maybe we can also track the kind of IPv6 connectivity (none, tunnelled, native) > Regards, > Robert > Thanks, Florian -- I remember yesterday, but the memory is in my head now. Was yesterday real? Or is it only the memory that is real? From dk at openit.de Thu Jan 13 10:51:08 2011 From: dk at openit.de (Dennis Koegel) Date: Thu, 13 Jan 2011 10:51:08 +0100 Subject: [atlas]Some new UI features In-Reply-To: <4D2EC360.7060903@narrans.de> References: <4D2D72D8.9040901@ripe.net> <4D2EC360.7060903@narrans.de> Message-ID: <20110113095108.GA8221@openit.de> Hi, I'd like to suggest the AS number for the list columns (could be determined automatically), and also a search box (id, description, location, asn) might come in handy if the list of public probes continues to grow. I guess sorting by geo /distance/ to the viewer's location (which could be manually configured or just set to some probe's location, e.g. an own probe) would be very helpful as well, as I often like to compare how things look like from "somewhere completely else". Thanks! - D. From patrick at vande-walle.eu Mon Jan 31 09:30:41 2011 From: patrick at vande-walle.eu (Patrick Vande Walle) Date: Mon, 31 Jan 2011 09:30:41 +0100 Subject: [atlas]Strange behaviour of probe Message-ID: <2e4da5863b44428a06ed0b1c8cca891d@mail.isoc.lu> Hello Robert, According to the graphs on atlas.ripe.net, my probe (#45) seems to have been offline for nearly 24 hours. During that period, the DSL connection was up nevertheless. Are there more detailed logs on the probe controller that could explain what happened ? Thanks, Patrick Vande Walle From daniel.karrenberg at ripe.net Mon Jan 31 10:04:32 2011 From: daniel.karrenberg at ripe.net (Daniel Karrenberg) Date: Mon, 31 Jan 2011 10:04:32 +0100 Subject: [atlas]Strange behaviour of probe In-Reply-To: <2e4da5863b44428a06ed0b1c8cca891d@mail.isoc.lu> References: <2e4da5863b44428a06ed0b1c8cca891d@mail.isoc.lu> Message-ID: <20110131090432.GB1056@reif.karrenberg.net> On 31.01 09:30, Patrick Vande Walle wrote: > Hello Robert, > > According to the graphs on atlas.ripe.net, my probe (#45) seems to have > been offline for nearly 24 hours. > During that period, the DSL connection was up nevertheless. > > Are there more detailed logs on the probe controller that could explain > what happened ? Just a quick response: I am aware that a number of probes have developed problems with time keeping due to one of our time servers dying. This problem is of the type 'should not happen' and we are analysing why it defies our mental model about the firmware ;-). This may well be related and affect more probes. Please bear with us while we deal with it. We will keep this list informed about progress. Daniel