From bortzmeyer at nic.fr Mon Jan 7 21:31:51 2013 From: bortzmeyer at nic.fr (Stephane Bortzmeyer) Date: Mon, 7 Jan 2013 21:31:51 +0100 Subject: [atlas] [UDM] Cannot select "recursion desired" for DNS Message-ID: <20130107203151.GA26749@sources.org> When creating a new UDM with DNS (or DNS6), the checkbox "Recursion desired" is greyed and un-selectable. I know that recursion can request more resources from the DNS servers. Is it why it is currently unavailable? It does not seem documented in which just says: Recursion desired Enable recursion. This is the RD flag described in RFC1035. From olaf.maunz at me.com Mon Jan 7 22:57:44 2013 From: olaf.maunz at me.com (Olaf Maunz) Date: Mon, 07 Jan 2013 21:57:44 +0000 Subject: [atlas] UI graphs Message-ID: The graphics for the pings seem to have vanished in the UI. At least for my account... From f.bauhaus at portrix-systems.de Tue Jan 8 08:15:44 2013 From: f.bauhaus at portrix-systems.de (Florian Bauhaus) Date: Tue, 08 Jan 2013 08:15:44 +0100 Subject: [atlas] UI graphs In-Reply-To: References: Message-ID: <50EBC7A0.8030804@portrix-systems.de> On 01/07/13 22:57, Olaf Maunz wrote: > The graphics for the pings seem to have vanished in the UI. At least for my account... > If you open the ping UDM there is (at least for me) an tab that says 'RRDs' and there you have all the graphs. Florian -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 894 bytes Desc: OpenPGP digital signature URL: From antony at ripe.net Tue Jan 8 10:11:25 2013 From: antony at ripe.net (Antony Antony) Date: Tue, 8 Jan 2013 10:11:25 +0100 Subject: [atlas] [UDM] Cannot select "recursion desired" for DNS In-Reply-To: <20130107203151.GA26749@sources.org> References: <20130107203151.GA26749@sources.org> Message-ID: <20130108091125.GA8456@dog.ripe.net> It will be enabled in the next release, I think before end fo the month. Initially we were waiting for majority probes to have the supported firmware. However, by default RD is enabled when querying probe's local resolver. Also my experience is som resolvers ignore it; regards, -antony On Mon, Jan 07, 2013 at 09:31:51PM +0100, Stephane Bortzmeyer wrote: > When creating a new UDM with DNS (or DNS6), the checkbox "Recursion > desired" is greyed and un-selectable. > > I know that recursion can request more resources from the DNS > servers. Is it why it is currently unavailable? > > It does not seem documented in which > just says: > > Recursion desired > Enable recursion. This is the RD flag described in RFC1035. > > > From bortzmeyer at nic.fr Tue Jan 8 13:03:51 2013 From: bortzmeyer at nic.fr (Stephane Bortzmeyer) Date: Tue, 8 Jan 2013 13:03:51 +0100 Subject: [atlas] [UDM] Cannot select "recursion desired" for DNS In-Reply-To: <20130108091125.GA8456@dog.ripe.net> References: <20130107203151.GA26749@sources.org> <20130108091125.GA8456@dog.ripe.net> Message-ID: <20130108120351.GA27969@nic.fr> On Tue, Jan 08, 2013 at 10:11:25AM +0100, Antony Antony wrote a message of 26 lines which said: > It will be enabled in the next release, OK, thanks. > However, by default RD is enabled when querying probe's local > resolver. Yes, but I wanted to test public resolvers. > Also my experience is som resolvers ignore it; My experience with public resolvers does not match yours: Google Public DNS: without RD, serves only the data in the cache (otherwise, if not in cache, NOERROR,ANSWER=0) OpenDNS: without RD, always REFUSED, even if data in the cache Level 3: like OpenDNS Comodo: like Level 3 and OpenDNS DNS Advantage: without RD, always timeout, even if data in the cache Telecomix: like DNS Advantage Regular Unbound: without RD, serves only the data in the cache (otherwise, if not in cache, REFUSED) Regular BIND: without RD, serves only the data in the cache (otherwise, if not in cache, NOERROR,ANSWER=0 but with upward referral) From pk at DENIC.DE Tue Jan 8 14:08:57 2013 From: pk at DENIC.DE (Peter Koch) Date: Tue, 8 Jan 2013 14:08:57 +0100 Subject: [atlas] [UDM] Cannot select "recursion desired" for DNS In-Reply-To: <20130108091125.GA8456@dog.ripe.net> References: <20130107203151.GA26749@sources.org> <20130108091125.GA8456@dog.ripe.net> Message-ID: <20130108130857.GA11174@x28.adm.denic.de> On Tue, Jan 08, 2013 at 10:11:25AM +0100, Antony Antony wrote: > However, by default RD is enabled when querying probe's local resolver. attention, significant paranoia ahead, independent of RD: Are there any restrictions on QNAMEs that can be sent to the local resolver or could a UDM be used to do reconnaissance against the "surrounding" namespace (or address space wit the reverse tree)? What do the probes do if DHCP does not provide for local resolvers? -Peter From philip.homburg at ripe.net Tue Jan 8 15:22:43 2013 From: philip.homburg at ripe.net (Philip Homburg) Date: Tue, 08 Jan 2013 15:22:43 +0100 Subject: [atlas] [UDM] Cannot select "recursion desired" for DNS In-Reply-To: <20130108130857.GA11174@x28.adm.denic.de> References: <20130107203151.GA26749@sources.org> <20130108091125.GA8456@dog.ripe.net> <20130108130857.GA11174@x28.adm.denic.de> Message-ID: <50EC2BB3.5030200@ripe.net> On 1/8/13 14:08 , Peter Koch wrote: > On Tue, Jan 08, 2013 at 10:11:25AM +0100, Antony Antony wrote: > >> However, by default RD is enabled when querying probe's local resolver. > attention, significant paranoia ahead, independent of RD: > > Are there any restrictions on QNAMEs that can be sent to the local resolver > or could a UDM be used to do reconnaissance against the "surrounding" > namespace (or address space wit the reverse tree)? There are no restrictions on what can be queried. > What do the probes do if DHCP does not provide for local resolvers? > > Probes can connect to the registration server without getting a resolver from DHCP. Beyond that, probes need a resolver. But you can, for example, configure a public DNS resolver statically (though there is no reason not to do that through DHCP). From nrw at kaffeeschluerfer.com Thu Jan 10 11:53:28 2013 From: nrw at kaffeeschluerfer.com (nrw at kaffeeschluerfer.com) Date: Thu, 10 Jan 2013 11:53:28 +0100 Subject: [atlas] UI graphs Message-ID: <20130110105328.276640@gmx.net> On Tue Jan 8 08:15:44 CET 2013, Florian Bauhaus f.bauhaus at portrix-systems.de wrote: > On 01/07/13 22:57, Olaf Maunz wrote: > > The graphics for the pings seem to have vanished in the UI. At least for my account... > > > > If you open the ping UDM there is (at least for me) an tab that says 'RRDs' and there you have all the graphs. I think Olaf did not refer to the UI of the UDMs, but to the UI of the probes. At least for me as well, the graphs of the statistics started to show no data for a period starting sometime between Jan 5 and Jan 6, 2013. At some point, I got the message that there was some maintenance going on on the Atlas infrastructure side. However, nothing changed with respect to the graphs not showing current data once the maintenance was through. Interestingly, the graphs started showing something again for data collected starting as of about 8 AM UTC this morning. Apart from some interruptions due to my public IP address changing, my probe was considered "up" all the time, and I also saw it running its tests and getting responses to the various probe messages. However I still need to figure out (I am sure it is in the documentation, I just didn't get around to looking for the relevant parts yet) how to access the raw data for recent measurements to see whether that data not being displayed in the graphs was due to the collected data somehow not getting to the Atlas infrastructure for some reason (despite probe being considered up), or whether there was just an issue displaying the collected data. Has anyone else seen similar issues and/or some idea what might have happened? Thanks! Kind regards, Joachim From nrw at kaffeeschluerfer.com Thu Jan 10 12:15:25 2013 From: nrw at kaffeeschluerfer.com (nrw at kaffeeschluerfer.com) Date: Thu, 10 Jan 2013 12:15:25 +0100 Subject: [atlas] UI graphs In-Reply-To: <20130110105328.276640@gmx.net> References: <20130110105328.276640@gmx.net> Message-ID: <20130110111525.276640@gmx.net> > However I still need to figure out (I am sure it is in the documentation, > I just didn't get around to looking for the relevant parts yet) how to > access the raw data for recent measurements to see whether that data not being > displayed in the graphs was due to the collected data somehow not getting > to the Atlas infrastructure for some reason (despite probe being considered > up), or whether there was just an issue displaying the collected data. Ok, getting the current raw data was of course much easier than I thought: No need to study the documentation, just to have a fresh look at the probe UI :-) Now with a first quick glance at the data of one built-in measurement, it seems that data was received by the Atlas infrastructure e.g. for Tue, 8 Jan 2013 20:17:43 UTC. Still need to check further whether there is also data for the other measurements, and whether that data is valid. But it seems there was just a glitch in rendering the data. Kind regards, Joachim From astrikos at ripe.net Thu Jan 10 12:28:40 2013 From: astrikos at ripe.net (Andreas Strikos) Date: Thu, 10 Jan 2013 12:28:40 +0100 Subject: [atlas] UI graphs In-Reply-To: <20130110105328.276640@gmx.net> References: <20130110105328.276640@gmx.net> Message-ID: <50EEA5E8.9070905@ripe.net> Hi all, Indeed one of the machines that is serving the graphs for probes between 2000-2499 had two issues. First one started somewhere on mid of 5th of January, as Joachim pointed out, and the next one was on 8th. The result was to miss the updates for the RRDs. Raw data though is still there, so theoretically we could recreate the RRDs if that's what you and the rest hosts of these probes want. But in this case we will have to suspend the update of the RRDs again for several hours. Sorry for any inconvenience. Regards, Andreas On 1/10/13 11:53 AM, nrw at kaffeeschluerfer.com wrote: > On Tue Jan 8 08:15:44 CET 2013, Florian Bauhaus f.bauhaus at portrix-systems.de wrote: >> On 01/07/13 22:57, Olaf Maunz wrote: >>> The graphics for the pings seem to have vanished in the UI. At least for my account... >>> >> If you open the ping UDM there is (at least for me) an tab that says > 'RRDs' and there you have all the graphs. > > I think Olaf did not refer to the UI of the UDMs, but to the UI of the probes. At least for me as well, the graphs of the statistics started to show no data for a period starting sometime between Jan 5 and Jan 6, 2013. > > At some point, I got the message that there was some maintenance going on on the Atlas infrastructure side. However, nothing changed with respect to the graphs not showing current data once the maintenance was through. > > Interestingly, the graphs started showing something again for data collected starting as of about 8 AM UTC this morning. > > Apart from some interruptions due to my public IP address changing, my probe was considered "up" all the time, and I also saw it running its tests and getting responses to the various probe messages. > > However I still need to figure out (I am sure it is in the documentation, I just didn't get around to looking for the relevant parts yet) how to access the raw data for recent measurements to see whether that data not being displayed in the graphs was due to the collected data somehow not getting to the Atlas infrastructure for some reason (despite probe being considered up), or whether there was just an issue displaying the collected data. > > Has anyone else seen similar issues and/or some idea what might have happened? > > Thanks! > > Kind regards, > > Joachim > From nrw at kaffeeschluerfer.com Thu Jan 10 15:35:06 2013 From: nrw at kaffeeschluerfer.com (nrw at kaffeeschluerfer.com) Date: Thu, 10 Jan 2013 15:35:06 +0100 Subject: [atlas] UI graphs In-Reply-To: <50EEA5E8.9070905@ripe.net> References: <20130110105328.276640@gmx.net> <50EEA5E8.9070905@ripe.net> Message-ID: <20130110143506.35830@gmx.net> Hi Andreas, Thanks for this explanation. At least from my side, there is no need to recreate those graphs. I was just a bit confused and didn't know if there was anything wrong with the network setup on my side or the probe, also not seeing any mentioning of this (apparently known issue) on the mailing list before Olaf's post. Kind regards, Joachim -------- Original-Nachricht -------- > Datum: Thu, 10 Jan 2013 12:28:40 +0100 > Von: Andreas Strikos > An: ripe-atlas at ripe.net > CC: nrw at kaffeeschluerfer.com > Betreff: Re: [atlas] UI graphs > Hi all, > > Indeed one of the machines that is serving the graphs for probes between > 2000-2499 had two issues. > First one started somewhere on mid of 5th of January, as Joachim pointed > out, and the next one was on 8th. The result was to miss the updates for > the RRDs. > Raw data though is still there, so theoretically we could recreate the > RRDs if that's what you and the rest hosts of these probes want. > But in this case we will have to suspend the update of the RRDs again > for several hours. > > Sorry for any inconvenience. > > Regards, > Andreas > > On 1/10/13 11:53 AM, nrw at kaffeeschluerfer.com wrote: > > On Tue Jan 8 08:15:44 CET 2013, Florian Bauhaus f.bauhaus at > portrix-systems.de wrote: > >> On 01/07/13 22:57, Olaf Maunz wrote: > >>> The graphics for the pings seem to have vanished in the UI. At least > for my account... > >>> > >> If you open the ping UDM there is (at least for me) an tab that says > > 'RRDs' and there you have all the graphs. > > > > I think Olaf did not refer to the UI of the UDMs, but to the UI of the > probes. At least for me as well, the graphs of the statistics started to > show no data for a period starting sometime between Jan 5 and Jan 6, 2013. > > > > At some point, I got the message that there was some maintenance going > on on the Atlas infrastructure side. However, nothing changed with respect > to the graphs not showing current data once the maintenance was through. > > > > Interestingly, the graphs started showing something again for data > collected starting as of about 8 AM UTC this morning. > > > > Apart from some interruptions due to my public IP address changing, my > probe was considered "up" all the time, and I also saw it running its tests > and getting responses to the various probe messages. > > > > However I still need to figure out (I am sure it is in the > documentation, I just didn't get around to looking for the relevant parts yet) how to > access the raw data for recent measurements to see whether that data not > being displayed in the graphs was due to the collected data somehow not getting > to the Atlas infrastructure for some reason (despite probe being > considered up), or whether there was just an issue displaying the collected data. > > > > Has anyone else seen similar issues and/or some idea what might have > happened? > > > > Thanks! > > > > Kind regards, > > > > Joachim > > > From bortzmeyer at nic.fr Tue Jan 15 10:14:13 2013 From: bortzmeyer at nic.fr (Stephane Bortzmeyer) Date: Tue, 15 Jan 2013 10:14:13 +0100 Subject: [atlas] Parametrization of UDM requests? Message-ID: <20130115091413.GA27606@nic.fr> It seems that the parameters used for an UDM are the same for all probes. For instance, I would like to do a DNS UDM with a different question for each probe, something like "%p.mydomain.example" where %p would be replaced by the Probe ID. Is it possible? Or do you think it would be a worthwile improvment? From bortzmeyer at nic.fr Tue Jan 15 11:50:17 2013 From: bortzmeyer at nic.fr (Stephane Bortzmeyer) Date: Tue, 15 Jan 2013 11:50:17 +0100 Subject: [atlas] Parametrization of UDM requests? In-Reply-To: <20130115091413.GA27606@nic.fr> References: <20130115091413.GA27606@nic.fr> Message-ID: <20130115105017.GA5783@nic.fr> On Tue, Jan 15, 2013 at 10:14:13AM +0100, Stephane Bortzmeyer wrote a message of 6 lines which said: > It seems that the parameters used for an UDM are the same for all > probes. For instance, I would like to do a DNS UDM with a different > question for each probe, something like "%p.mydomain.example" where > %p would be replaced by the Probe ID. Another solution would be of course an API to run measurements. Then, I would develop the feature, not the Atlas team :-) From robert at ripe.net Tue Jan 15 12:10:03 2013 From: robert at ripe.net (Robert Kisteleki) Date: Tue, 15 Jan 2013 12:10:03 +0100 Subject: [atlas] Parametrization of UDM requests? In-Reply-To: <20130115105017.GA5783@nic.fr> References: <20130115091413.GA27606@nic.fr> <20130115105017.GA5783@nic.fr> Message-ID: <50F5390B.8050206@ripe.net> Hi Stephane, On 2013.01.15. 11:50, Stephane Bortzmeyer wrote: > On Tue, Jan 15, 2013 at 10:14:13AM +0100, > Stephane Bortzmeyer wrote > a message of 6 lines which said: > >> It seems that the parameters used for an UDM are the same for all >> probes. For instance, I would like to do a DNS UDM with a different >> question for each probe, something like "%p.mydomain.example" where >> %p would be replaced by the Probe ID. Thank you for the suggestion! We're investigating this, but in the meantime can you explain what particular problem would this solve? > Another solution would be of course an API to run measurements. Then, > I would develop the feature, not the Atlas team :-) I have good news for you: we have such an API, and we'll document and publish it soon. Regards, Robert From bortzmeyer at nic.fr Tue Jan 15 12:43:00 2013 From: bortzmeyer at nic.fr (Stephane Bortzmeyer) Date: Tue, 15 Jan 2013 12:43:00 +0100 Subject: [atlas] Parametrization of UDM requests? In-Reply-To: <50F5390B.8050206@ripe.net> References: <20130115091413.GA27606@nic.fr> <20130115105017.GA5783@nic.fr> <50F5390B.8050206@ripe.net> Message-ID: <20130115114300.GA11586@nic.fr> On Tue, Jan 15, 2013 at 12:10:03PM +0100, Robert Kisteleki wrote a message of 23 lines which said: > We're investigating this, but in the meantime can you explain what > particular problem would this solve? I want to investigate public anycast resolvers like Google Public DNS. I send a query to 8.8.8.8 for a domain I control. On the authoritative name server, I see the requests, I see the IP address of the resolver, I don't know which probe was responsible since they all use the same question (QNAME, in DNS parlance). From BECHA at ripe.net Wed Jan 16 15:59:42 2013 From: BECHA at ripe.net (Vesna Manojlovic) Date: Wed, 16 Jan 2013 15:59:42 +0100 Subject: [atlas] Thank you for making the 2012 a great year for RIPE Atlas! Message-ID: <50F6C05E.9000408@ripe.net> Dear RIPE Atlas community, many thanks for hosting, testing & using RIPE Atlas! We have reached many milestones last year, and are looking forward to new challenges this year. Please see the story of the RIPE Atlas achievements in this RIPE Labs article: https://labs.ripe.net/Members/becha/ripe-atlas-2012-a-year-in-review Regards, Vesna -- Vesna Manojlovic BECHA at ripe.net Senior Community Builder @Ms_Measurements for Measurements Tools RIPE NCC http://ripe.net From jprins at betterbe.com Wed Jan 23 09:03:30 2013 From: jprins at betterbe.com (Jan Hugo Prins) Date: Wed, 23 Jan 2013 09:03:30 +0100 Subject: [atlas] 2 probes down Message-ID: <50FF9952.8050205@betterbe.com> Hello, This morning I got a notification that the probe in my datacentre is down. I checked everything and he looks just fine and is happily doing traffic. Then I noticed that the one behind my home fiber connection is also down, and that the starttime is about the same. Both of them at 6:56 this morning. Also checked that one and he is also doing traffic just fine. Houston, do we hava a problem? -- Met vriendelijke groet / Best regards, Jan Hugo Prins Infra and Isilon storage consultant E: jprins at betterbe.com T: +31-53-4800694 M: +31-6-26358951 S: jhaprins W: www.betterbe.com -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: EMC-Isilon-ICIE-sm.jpg Type: image/jpeg Size: 4558 bytes Desc: not available URL: From robert at ripe.net Wed Jan 23 10:08:37 2013 From: robert at ripe.net (Robert Kisteleki) Date: Wed, 23 Jan 2013 10:08:37 +0100 Subject: [atlas] 2 probes down In-Reply-To: <50FF9952.8050205@betterbe.com> References: <50FF9952.8050205@betterbe.com> Message-ID: <50FFA895.30202@ripe.net> Dear Jan Hugo, Indeed, around 06:50 UTC two of our controllers became unreachable. ("Controllers" are the machines that probes are continuously connected to.) They are in the same colo (not at the RIPE NCC), both were up and running otherwise, and no other servers were affected so we can safely assume that the colo lost its connectivity briefly. In these cases the probes will reconnect after a short time (your two probes are back too). Regards, Robert RIPE Atlas team On 2013.01.23. 9:03, Jan Hugo Prins wrote: > Hello, > > This morning I got a notification that the probe in my datacentre is down. I > checked everything and he looks just fine and is happily doing traffic. Then > I noticed that the one behind my home fiber connection is also down, and > that the starttime is about the same. Both of them at 6:56 this morning. > Also checked that one and he is also doing traffic just fine. > > Houston, do we hava a problem? > > -- > Met vriendelijke groet / Best regards, > > Jan Hugo Prins > Infra and Isilon storage consultant > > > E: jprins at betterbe.com > T: +31-53-4800694 > M: +31-6-26358951 > S: jhaprins > W: www.betterbe.com -------------- next part -------------- A non-text attachment was scrubbed... Name: tealc_last_10800.png Type: image/png Size: 27896 bytes Desc: not available URL: