From purri at openmailbox.org Wed Jan 4 16:01:34 2017 From: purri at openmailbox.org (purri) Date: Wed, 4 Jan 2017 19:01:34 +0300 Subject: [dnstap] Knot+Dnstap how to read log in realtime? Message-ID: <8437bf7c-9e9b-e660-e8d1-6987ac873f87@openmailbox.org> I tried to use knot-2.3.3 with dnstap hen i use "sink: /tmp/capture.tap" I ran into a problem that capture.tap updates only every 4096 bytes I also tried this command to create socket: "fstrm_capture -t protobuf:dnstap.Dnstap -u /tmp/dnstap.sock -w /tmp/capture.tap -ddddd" with "sink: /tmp/dnstap.sock" but here capture.tap never filled, I restarted knotd , but looks like it not even connect to this sock. From each at isc.org Wed Jan 4 17:31:24 2017 From: each at isc.org (Evan Hunt) Date: Wed, 4 Jan 2017 17:31:24 +0000 Subject: [dnstap] Knot+Dnstap how to read log in realtime? In-Reply-To: <8437bf7c-9e9b-e660-e8d1-6987ac873f87@openmailbox.org> References: <8437bf7c-9e9b-e660-e8d1-6987ac873f87@openmailbox.org> Message-ID: <20170104173124.GA8453@isc.org> On Wed, Jan 04, 2017 at 07:01:34PM +0300, purri wrote: > I tried to use knot-2.3.3 with dnstap > > hen i use "sink: /tmp/capture.tap" I ran into a problem that capture.tap > updates only every 4096 bytes > > I also tried this command to create socket: > "fstrm_capture -t protobuf:dnstap.Dnstap -u /tmp/dnstap.sock -w > /tmp/capture.tap -ddddd" > with "sink: /tmp/dnstap.sock" but here capture.tap never filled, > I restarted knotd , but looks like it not even connect to this sock. I remember submitting a change to Robert so that a SIGHUP to fstrm_capture would cause it to flush its output immediately. I'm not sure whether he ever accepted the patch though. -- Evan Hunt -- each at isc.org Internet Systems Consortium, Inc. From edmonds at mycre.ws Wed Jan 4 17:45:21 2017 From: edmonds at mycre.ws (Robert Edmonds) Date: Wed, 4 Jan 2017 12:45:21 -0500 Subject: [dnstap] Knot+Dnstap how to read log in realtime? In-Reply-To: <20170104173124.GA8453@isc.org> References: <8437bf7c-9e9b-e660-e8d1-6987ac873f87@openmailbox.org> <20170104173124.GA8453@isc.org> Message-ID: <20170104174521.rfy6r6u5ji5ea2ku@mycre.ws> Evan Hunt wrote: > I remember submitting a change to Robert so that a SIGHUP to > fstrm_capture would cause it to flush its output immediately. I'm > not sure whether he ever accepted the patch though. It was added in the fstrm 0.3.0 release: https://github.com/farsightsec/fstrm/pull/12 -- Robert Edmonds