Mailing-List: contact general-help@incubator.apache.org; run by ezmlm
Precedence: bulk
Reply-To: general@incubator.apache.org
MIME-Version: 1.0
In-Reply-To: <E0217E74-6149-4A00-8962-70956DD3102C@cisco.com>
References: <CAOENjmi1HJ=EPxxkyjpfH1Pv51EPs8KS2Jy4buX_HBHznXeCiA@mail.gmail.com>
 <3B954DCC-69FD-4210-9D72-5EB2D4E22575@cisco.com> <5E326869-544A-45B9-AB99-EDAA5BB44E52@classsoftware.com>
 <7A1D486C-7703-4280-8F82-4930D34BB076@classsoftware.com> <CAOqetn_+u2fWrRX721v9_AqqY3hdfmEG6G_2jB7H58_YCP8wZA@mail.gmail.com>
 <CAHkg7wTmsqsdcdaVYdmZiwEuTQBZdfPk-32mpeEbGzNz243NkA@mail.gmail.com>
 <5127D6FC-D379-4C3E-A1C2-99BAE614DFD7@classsoftware.com> <CABF6JR3_fZgd1z6X-rx8kGOMQvmuQcK-zKi=DXY=wBxMw9VO+w@mail.gmail.com>
 <E0217E74-6149-4A00-8962-70956DD3102C@cisco.com>
From: Dave Neuman <neuman@apache.org>
Date: Mon, 20 Nov 2017 18:13:55 +0000
Message-ID: <CAMuiy1PvTYuw2AV+k+FH=tHmGt2skCCghZcni53iFP5K-=mZ3w@mail.gmail.com>
Subject: Re: [RESULT] [VOTE] Release Apache Traffic Control 2.0.0 incubating (RC6)
To: general@incubator.apache.org
Content-Type: multipart/alternative; boundary="f40304397c84b9c767055e6e0cdb"
archived-at: Mon, 20 Nov 2017 18:14:01 -0000

--f40304397c84b9c767055e6e0cdb
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

+1 However....
We need to add a release note suggesting that traffic stats works best with
influxdb version < 1.3.x.  As of 1.3.x InfluxDB now returns a 400 response
when the client attempts to write points that are outside of the retention
policy.  When this happens, Traffic Stats seems to hold on to the "old"
points and attempts to write them again on the next POST.  This causes what
is essentially a memory leak in Traffic Stats since it continues to hold
onto and tries to write stats that are outside of the retention policy.
This may or may not affect a user, depending upon which stats they are
trying to write to influxdb.  For example, we write the wrap_count to
influxdb, this is a stat that does not often change within 24 hours, so we
see the memory leak with stats not written on each poll of traffic stats.

In versions < 1.3.x InfluxDB would still not write the point, but it would
accept the write request and just drop the points outside of the retention
policy on the floor.

This is fixed here:
https://github.com/apache/incubator-trafficcontrol/pull/1289 and will be in
the next release of traffic control.

Also, we should note that you need to set selinux to "unenforcing" in order
to get the build to work, otherwise you fail on permissions erros.  You can
do this with
`setenforce 0` as root.


Thanks,
Dave

On Thu, Jul 6, 2017 at 3:26 PM, Eric Friedrich (efriedri) <
efriedri@cisco.com> wrote:

> Dear IPMC Community,
>
> I am pleased to announce that the Incubator PMC has approved 2.0.0-RC6 of
> Apache Traffic Control (Incubating) for release as =E2=80=9CApache Traffi=
c Control
> (incubating) 2.0.0"
>
> The vote has passed with:
> - three binding "+1" votes
> - no "0" votes
> - no "-1" votes
>
> The votes were (https://lists.apache.org/thread.html/
> c44716e02d175b78def5d1a219474e6f5c397120e33d64208158511e@%
> 3Cgeneral.incubator.apache.org%3E<https://lists.apache.org/thread.html/
> c44716e02d175b78def5d1a219474e6f5c397120e33d64208158511e@<ge
> neral.incubator.apache.org>>):
> - +1, Justin Mclean (binding)
> - +1, John D. Ament (binding)
> - +1, Phil Sorber (binding)
>
> Thank you, for your support!
>
> =E2=80=94Eric
>
> On Jun 30, 2017, at 5:27 PM, Phil Sorber <sorber@apache.org<mailto:sorb
> er@apache.org>> wrote:
>
> Checksums match.
>
> GPG sig verifies, but is untrusted:
> $ gpg --verify apache-trafficcontrol-2.0.0-incubating.tar.gz.asc
> gpg: assuming signed data in `apache-trafficcontrol-2.0.0-
> incubating.tar.gz'
> gpg: Signature made Fri 16 Jun 2017 10:25:33 AM MDT using RSA key ID
> 9AB7BDD5
> gpg: Good signature from "Eric Friedrich (CODE SIGNING KEY) <
> friede@apache.org<mailto:friede@apache.org>>"
> gpg: WARNING: This key is not certified with a trusted signature!
> gpg:          There is no indication that the signature belongs to the
> owner.
> Primary key fingerprint: C3ED F3D4 C5EE F3E5 5932  3FF8 F220 0BAB 9AB7 BD=
D5
>
> No signatures:
> $ gpg --list-sigs 9AB7BDD5
> pub   4096R/9AB7BDD5 2017-04-25
> uid                  Eric Friedrich (CODE SIGNING KEY) <friede@apache.org
> <mailto:friede@apache.org>>
> sig 3        9AB7BDD5 2017-04-25  Eric Friedrich (CODE SIGNING KEY) <
> friede@apache.org<mailto:friede@apache.org>>
> sub   4096R/F445ABB1 2017-04-25
> sig          9AB7BDD5 2017-04-25  Eric Friedrich (CODE SIGNING KEY) <
> friede@apache.org<mailto:friede@apache.org>>
>
> I won't -1 on that, but we should try to have every RM make sure their ke=
ys
> are trusted. I've already spoken with Eric about this.
>
> Build also works, with one caveat. You have to set selinux to unenforcing=
,
> otherwise you fail from permissions errors. You can do this with
> `setenforce 0` as root. I won't -1 for this either, but we should add tha=
t
> to the build documentation.
>
> -rw-r--r--. 1 phil phil  51M Jun 30 15:12
> traffic_monitor-2.0.0-5668.85d14ebe.el7.x86_64.rpm
> -rw-r--r--. 1 phil phil 910K Jun 30 15:10
> traffic_ops-2.0.0-5668.85d14ebe.el7.src.rpm
> -rw-r--r--. 1 phil phil 858K Jun 30 15:10
> traffic_ops-2.0.0-5668.85d14ebe.el7.x86_64.rpm
> -rw-r--r--. 1 phil phil  22K Jun 30 15:10
> traffic_ops_ort-2.0.0-5668.85d14ebe.el7.src.rpm
> -rw-r--r--. 1 phil phil  21K Jun 30 15:10
> traffic_ops_ort-2.0.0-5668.85d14ebe.el7.x86_64.rpm
> -rw-r--r--. 1 phil phil 142K Jun 30 15:12
> traffic_portal-2.0.0-5668.85d14ebe.el7.src.rpm
> -rw-r--r--. 1 phil phil 3.6M Jun 30 15:12
> traffic_portal-2.0.0-5668.85d14ebe.el7.x86_64.rpm
> -rw-r--r--. 1 phil phil  17M Jun 30 15:12
> traffic_router-2.0.0-5668.85d14ebe.el7.x86_64.rpm
> -rw-r--r--. 1 phil phil 417K Jun 30 15:11
> traffic_stats-2.0.0-5668.85d14ebe.el7.src.rpm
> -rw-r--r--. 1 phil phil 5.0M Jun 30 15:11
> traffic_stats-2.0.0-5668.85d14ebe.el7.x86_64.rpm
>
> +1 binding.
>
> Thanks!
>
> On Wed, Jun 28, 2017 at 5:20 PM Justin Mclean <justin@classsoftware.com<
> mailto:justin@classsoftware.com>>
> wrote:
>
> Hi,
>
> The build instructions will be put in place once the release is done
> -- catch-22.
>
> Yep another reason for including them in the release IMO.
>
> Would Jira issues for the copyright and build instruction issues be
> sufficient?
>
> Yes. Changing my vote to +1 binding.
>
> Thanks,
> Justin
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org<mailto:
> general-unsubscribe@incubator.apache.org>
> For additional commands, e-mail: general-help@incubator.apache.org<mailto=
:
> general-help@incubator.apache.org>
>
>
>
>

--f40304397c84b9c767055e6e0cdb--