Notify time setting does not work
Probes have very handy feature of sending a notification when the probe is disconnected. I have mine set to 10 minutes, however last two such notifications only arrived 30 minutes after the probe was disconnected. Probe #26656 if that's important. One would think that probe itself was slow to react, but last time I monitored outage from the very beginning and the probe's page correctly showed "Disconnected for: 1 minute" and so on. Neither it is a case of slow email arrival, since according to Received header in the email it was indeed received from worker1.atlas.ripe.net 30+ minutes after disconnection time mentioned in the email body. Can it be that "Notify time" setting just doesn't work and some default value of 30 minutes is used? -- With Best Regards, Marat Khalili
Hi Marat, On 17.08.2017 09:15, Marat Khalili wrote:
[...] Neither it is a case of slow email arrival, since according to Received header in the email it was indeed received from worker1.atlas.ripe.net 30+ minutes after disconnection time mentioned in the email body.
smells a bit as if greylisting would have been deployed for 'rqc.ru'. Cheers, -C.
smells a bit as if greylisting would have been deployed for 'rqc.ru'. You may be right, but why? It all happens still well within RIPE:
Received: from worker1.atlas.ripe.net ([193.0.19.216]) by mahimahi.ripe.net with esmtps (TLSv1.2:ECDHE-RSA-AES256-GCM-SHA384:256) (Exim 4.89) (envelope-from <no-reply@ripe.net>) id 1dhueG-0001AI-Sl for redacted@rqc.ru; Wed, 16 Aug 2017 11:30:02 +0200 Received: from [127.0.0.1] (helo=worker1.atlas.ripe.net) by worker1.atlas.ripe.net with esmtp (Exim 4.89) (envelope-from <no-reply@ripe.net>) id 1dhueG-0003zF-Qf for redacted@rqc.ru; Wed, 16 Aug 2017 09:30:00 +0000 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Subject: Probe 26656 is disconnected (RQC Ural bld.) From: RIPE Atlas (no reply) <no-reply@ripe.net> To: redacted@rqc.ru Reply-To: RIPE Atlas (no reply) <no-reply@ripe.net> Date: Wed, 16 Aug 2017 09:30:00 -0000 Message-ID: <20170816093000.18245.96974@worker1.atlas.ripe.net> X-ACL-Warn: Delaying message X-RIPE-Spam-Level: ---- X-RIPE-Spam-Report: Spam Total Points: -4.0 points pts rule name description ---- ---------------------- ------------------------------------ -7.5 ALL_TRUSTED Passed through trusted hosts only via SMTP 3.5 DCC_CHECK Detected as bulk mail by DCC (dcc-servers.net) X-RIPE-Signature: b7d297f3462a5391c8f90e217ab85ce69eb69a6fc89c21df31133e5b7354304c
Dear Marat Khalili,
Your probe 26656 has been disconnected from the RIPE Atlas infrastructure since 2017-08-16 08:58:54 UTC. [...] My provider did not treat it as spam and delivered almost instantly after receiving from RIPE.
-- With Best Regards, Marat Khalili
Hi Marat, On 17.08.2017 13:02, Marat Khalili wrote:
smells a bit as if greylisting would have been deployed for 'rqc.ru'. You may be right, but why? It all happens still well within RIPE:
nope - the devil, as always, is in the detail:
Received: from worker1.atlas.ripe.net ([193.0.19.216]) by mahimahi.ripe.net with esmtps (TLSv1.2:ECDHE-RSA-AES256-GCM-SHA384:256) (Exim 4.89) (envelope-from <no-reply@ripe.net>) id 1dhueG-0001AI-Sl for redacted@rqc.ru; Wed, 16 Aug 2017 11:30:02 +0200
11:30:02 +0200 is 09:30:02 +0000 (aka. UTC)
Received: from [127.0.0.1] (helo=worker1.atlas.ripe.net) by worker1.atlas.ripe.net with esmtp (Exim 4.89) (envelope-from <no-reply@ripe.net>) id 1dhueG-0003zF-Qf for redacted@rqc.ru; Wed, 16 Aug 2017 09:30:00 +0000
09:30:00 +0000 is 11:30:00 +0200
Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Subject: Probe 26656 is disconnected (RQC Ural bld.) From: RIPE Atlas (no reply) <no-reply@ripe.net> To: redacted@rqc.ru Reply-To: RIPE Atlas (no reply) <no-reply@ripe.net> Date: Wed, 16 Aug 2017 09:30:00 -0000
- so it took at least that very message "only" two seconds to hop from one NCC host to another NCC host: no "significant" delay here. Best, -C.
But: > Your probe 26656 has been disconnected from the RIPE Atlas > infrastructure since 2017-08-16 08:58:54 UTC What happened between 08:58:54 UTC and 09:30:00 +0000? -- With Best Regards, Marat Khalili On 17/08/17 14:17, Carsten Schiefner wrote: > Hi Marat, > > On 17.08.2017 13:02, Marat Khalili wrote: >>> smells a bit as if greylisting would have been deployed for 'rqc.ru'. >> You may be right, but why? It all happens still well within RIPE: > nope - the devil, as always, is in the detail: > >>> Received: from worker1.atlas.ripe.net ([193.0.19.216]) >>> by mahimahi.ripe.net with esmtps >>> (TLSv1.2:ECDHE-RSA-AES256-GCM-SHA384:256) >>> (Exim 4.89) >>> (envelope-from <no-reply@ripe.net>) >>> id 1dhueG-0001AI-Sl >>> for redacted@rqc.ru; Wed, 16 Aug 2017 11:30:02 +0200 > 11:30:02 +0200 is 09:30:02 +0000 (aka. UTC) > >>> Received: from [127.0.0.1] (helo=worker1.atlas.ripe.net) >>> by worker1.atlas.ripe.net with esmtp (Exim 4.89) >>> (envelope-from <no-reply@ripe.net>) >>> id 1dhueG-0003zF-Qf >>> for redacted@rqc.ru; Wed, 16 Aug 2017 09:30:00 +0000 > 09:30:00 +0000 is 11:30:00 +0200 > >>> Content-Type: text/plain; charset="utf-8" >>> MIME-Version: 1.0 >>> Content-Transfer-Encoding: 7bit >>> Subject: Probe 26656 is disconnected (RQC Ural bld.) >>> From: RIPE Atlas (no reply) <no-reply@ripe.net> >>> To: redacted@rqc.ru >>> Reply-To: RIPE Atlas (no reply) <no-reply@ripe.net> >>> Date: Wed, 16 Aug 2017 09:30:00 -0000 > - so it took at least that very message "only" two seconds to hop from > one NCC host to another NCC host: no "significant" delay here. > > Best, > > -C.
On 17.08.2017 13:34, Marat Khalili wrote:
But:
Your probe 26656 has been disconnected from the RIPE Atlas infrastructure since 2017-08-16 08:58:54 UTC What happened between 08:58:54 UTC and 09:30:00 +0000?
Erm, yes... Didn't check the mail body. Sorry, my bad. So it appears that this indeed *IS* onyl to the Atlas team @ NCC to answer. Cheers, -C.
Hi Marat, Here how it (always) works. There is a job started every 30 minutes. If a probe is down longer than the "Notify time" you will get a notification. So in your case (10 minutes) you should get an email within interval of [10 , 40) minutes. wbr /vty On 8/17/17 9:15 AM, Marat Khalili wrote:
Probes have very handy feature of sending a notification when the probe is disconnected. I have mine set to 10 minutes, however last two such notifications only arrived 30 minutes after the probe was disconnected. Probe #26656 if that's important.
One would think that probe itself was slow to react, but last time I monitored outage from the very beginning and the probe's page correctly showed "Disconnected for: 1 minute" and so on. Neither it is a case of slow email arrival, since according to Received header in the email it was indeed received from worker1.atlas.ripe.net 30+ minutes after disconnection time mentioned in the email body.
Can it be that "Notify time" setting just doesn't work and some default value of 30 minutes is used?
--
With Best Regards, Marat Khalili
Hi, On Fri, Aug 18, 2017 at 01:01:48PM +0200, Viktor Naumov wrote:
Here how it (always) works.
There is a job started every 30 minutes. If a probe is down longer than the "Notify time" you will get a notification. So in your case (10 minutes) you should get an email within interval of [10 , 40) minutes.
This is a bit awkward... can this be improved, like, run this job every minute or so? A single "select id from probes where downtime>notify and not is_notified" statement shouldn't cost much... (no ideas about your table setup, though :-) ) Gert Doering -- NetMaster -- have you enabled IPv6 on something today...? SpaceNet AG Vorstand: Sebastian v. Bomhard Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (0)89/32356-444 USt-IdNr.: DE813185279
participants (4)
-
Carsten Schiefner -
Gert Doering -
Marat Khalili -
Viktor Naumov