Age | Commit message (Collapse) | Author | Files | Lines |
|
Change-Id: I35ff6882b89191add29d700704fd235ed5fc4d27
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Change-Id: I115c83e231db7e52c322f2b5243ca84e423e9a54
Signed-off-by: Jan Srnicek <jsrnicek@cisco.com>
|
|
Change-Id: I52fe0e73dc6b0b66ae7d22991ea31a7210ee8d82
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Current configuration creates 4 plot sections,
but each of them contains only first of defined plots.
Change-Id: Ib8bcd2b3936ac08b9ca129b4e0902d1a3916ba5c
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
* Configure log shipping on all honeycomb jobs
Change-Id: Ib399a2e3d1da582d31a77927785f56f735684f0c
Signed-off-by: Vanessa Rene Valderrama <vvalderrama@linuxfoundation.org>
|
|
Depends on https://gerrit.fd.io/r/#/c/4385
Change-Id: Ib6f9e167c98d390248a9f2837b8bcb70a76af6ad
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
|
|
Transition jobs for honeycomb to be Packer compliant
* Modifying honeycomb build minions to use the new node naming standard
* Temporarily using mvn33-new during cutover
Change-Id: I359f2f6168235eabaacdfb2cb5183df8dd250f9b
Signed-off-by: Vanessa Rene Valderrama <vvalderrama@linuxfoundation.org>
|
|
Change-Id: I5b6085b0c675e66b14f5956f6526e68364a86da8
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Karaf is neither used in honeycomb nor in hc2vpp
Change-Id: I9f85f3391e5dd76b5d1e4b32d020da0ba3bf1b29
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
Change-Id: I3508a9528574d82ac813282aa2e03fc492503f5c
Signed-off-by: Maros Marsalek <mmarsale@cisco.com>
|
|
Change-Id: Ia974b445b506bd9259756ab5cb7663839d071c98
Signed-off-by: Maros Marsalek <mmarsale@cisco.com>
|
|
Change-Id: I44daa38007e4c5a4a7281b5034448f635726ac4c
Signed-off-by: Maros Marsalek <mmarsale@cisco.com>
|
|
|
|
|
|
|
|
|
|
Change-Id: I309ee62b065589e61a815e8f05e9cfc33742cc73
Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
|
|
- Move timestamps to fdio-infra-wrappers as every job is using it
- Move ssh-credentials into fdio-infra-wrappers as every job is using it
- Ensuer ALL jobs are using fdio-infra-wrappers (this should be policy!)
FD.io has this convenient wrapper called fdio-infra-wrappers which is a
simple macro to configure common wrapper functionality that all jobs
should be configured with all jobs are already configuring it too so
simplify job configuration and ensuring we don't have duplication
everywhere.
Less lines of code and more code reuse makes for easier maintenance :)
Change-Id: Ia73b09dfc556f684c43853797231da84a9c517af
Signed-off-by: Thanh Ha <thanh.ha@linuxfoundation.org>
|
|
ODL used to use a variable called {ssh-credentials} as a form of
redirection in order to have a simple human readable word to represent a
unique credentials id in Jenkins. Newer versions of Jenkins actually
allows us to override the credentials-id UUID format and pass a human
readable filename.
Considering that FD.io has been using a human readable
jenkins-gerrit-credentials credentials-id remove the unnecessary
{ssh-credentials} id defaults from global-defaults and just access the
credential directly.
Change-Id: I7b59357dcacb2195646719709b3b623429deca91
Signed-off-by: Thanh Ha <thanh.ha@linuxfoundation.org>
|
|
Instead of indirectly calling {mvn33} in jjb and redirecting it to
Maven_3.3.9 in Jenkins. Simply name Maven in Jenkins mvn33 and then
jobs can call mvn33 directly without the need for the defaults
redirection.
The new mvn33 profile has already been created in fd.io Jenkins.
Change-Id: I792509ab2d80546a532930f1f579afb69bfc670c
Signed-off-by: Thanh Ha <thanh.ha@linuxfoundation.org>
|
|
Change-Id: I8b976f59fe4a17ff0e620b17c40ba8f071a381ac
Signed-off-by: Maros Marsalek <mmarsale@cisco.com>
|
|
Jenkins 2 moved the JDK configuration to a new Jenkins global
configuration page under Manage Jenkins > Global Tool Configuration. The
fd.io instance of Jenkins was missing JDK configuration which as now
been rectified.
Additionally this patch configures the Jobs to run using a specified JDK
as configured via Jenkins global configuration. Valid values are:
- openjdk7-centos7
- openjdk8-centos8
- openjdk7-ubuntu1404
- openjdk8-ubuntu1404
- openjdk7-ubuntu1604
- openjdk8-ubuntu1604
This patch configures HoneyComb merge job to run using JDK8.
Follow on patches should be submitted to ensure all jobs explicitly
specify which JDK versions they need to run against.
RT: 30789
Change-Id: I71d281a0ea5b2dd3f1f45b9ea4acbff73765bd05
Signed-off-by: Thanh Ha <thanh.ha@linuxfoundation.org>
|
|
Sonar is failing at the moment because the build runs against Java 8 and
then sonar maven build step runs with Java 7. Disable sonar publishing
so that the merge job at least passes while we figure out the right
incantation to allow for the sonar step to pass.
Change-Id: I5c868ba3888a19aa85c46dc8ee6af4c6f2b49706
Signed-off-by: Thanh Ha <thanh.ha@linuxfoundation.org>
|
|
* prior to running the merge job, purge the openjdk-7-jre and openjdk-7-jdk packages
* sonar is failing due to not using the correct java version
* Specifying java 8 is insufficient
* Removing java 7 is the only working solution we've found.
Change-Id: I17d5d71c54f6fd3ea42afd8d96bb9281ba5cfce3
Signed-off-by: C.J. Collier <cjcollier@linuxfoundation.org>
|
|
* include jdk: openjdk8 as per discussion on IRC
Change-Id: I8cdedd5fed0129ff3ac6c56a98b235442d9ba3d3
Signed-off-by: C.J. Collier <cjcollier@linuxfoundation.org>
|
|
* honeycomb-merge-master-ubuntu1404 needs java 8 for sonar
* the revert to java 7 happens prior to the sonar run
* we now throw the guest away after each use, so there is no need to
revert to java 7
Change-Id: I551a5a6d3da5eb9d38cc1f624ef547815a41931c
Signed-off-by: C.J. Collier <cjcollier@linuxfoundation.org>
|
|
* honeycomb-merge-master-ubuntu1404 needs java 8 for sonar
* the revert to java 7 happens prior to the sonar run
* we now throw the guest away after each use, so there is no need to
revert to java 7
Change-Id: I82583ae4f6130961fb5c42741107ac74f80dd850
Signed-off-by: C.J. Collier <cjcollier@linuxfoundation.org>
|
|
To deploy site into docs.fd.io
Works on top of:
https://gerrit.fd.io/r/#/c/3073/
Site distribution configuration:
https://gerrit.fd.io/r/#/c/3073/5/common/honeycomb-parent/pom.xml
... only Ubuntu14 deploys the site during merge job
Note: using site instead of site:site, since only with site we
can hook in groovy maven plugin to fix some generated links. With
site:site, the groovy maven plugin does not get executed.
Change-Id: Ie1086b8d20856dbe15992e67ccc0d15547ce56d3
Signed-off-by: Maros Marsalek <mmarsale@cisco.com>
|
|
Change-Id: I3d246666a3180e35d787c0350755804531e6f1d2
Signed-off-by: Vanessa Rene Valderrama <vvalderrama@linuxfoundation.org>
|
|
No need, sonar analysis is run on merge jobs. No need to run again
from integration job. Also integration jobs might be run even if honeycomb
code hasen't changed, running the analysis pointlessly.
Change-Id: I527dbc77e9ce528da169443a77f47f5251216670
Signed-off-by: Maros Marsalek <mmarsale@cisco.com>
|
|
According to
http://stackoverflow.com/questions/15596522/sonar-analysis-using-multiple-git-branches-in-jenkins
this is the way of telling sonar to distinguish analysis for
multiple branches.
Change-Id: I599f69a3cc00d7c7541a182b0604bfb74df6b158
Signed-off-by: Maros Marsalek <mmarsale@cisco.com>
|
|
Triggers are succes on merge jobs for VPP or HC.
The template was bad though. It used {branch} instead
of {stream} in job name. But all jobs are defined using
{stream}. This works for master since master branch and stream
names are identical, but not for 1609 stable/1609.
Change-Id: Iaf7888bcfadb481fff51051eaee87e7613fd7a78
Signed-off-by: Maros Marsalek <mmarsale@cisco.com>
|
|
Change-Id: Iea3e23f53c327c35a50bb415a23dc733c0bada26
Signed-off-by: Maros Marsalek <mmarsale@cisco.com>
|
|
Change-Id: I9d7614eb4d0586ba42d1156d119c68e8bac285a1
Signed-off-by: Maros Marsalek <mmarsale@cisco.com>
|
|
* The script fails, though the failure can be ignored.
http://askubuntu.com/questions/599105/using-alternatives-with-java-7-and-java-8-on-14-04-2-lts
Change-Id: Ie7b70fb189505a46b6640f1a961001f0c75a21b5
Signed-off-by: C.J. Collier <cjcollier@linuxfoundation.org>
|
|
It turns out that doing rpm push in both merge and integration
job leads to collisions because they have two separate
sequences of BUILD_NUMBERs and those BUILD_NUMBERS are used
for the rpm release number.
So... this patch makes the merge job *stop* pushing rpms,
and should cause the integration job to trigger on both
vpp *and* honeycomb merge job success.
Change-Id: I12fe582ea8a1444096b86c214b3cd3c412bc7591
Signed-off-by: Ed Warnicke <eaw@cisco.com>
|
|
* remove deprecated MaxPermSize argument from mvn
* add repo-name parameter to job template definition
Change-Id: Ic202a8cbf56df52970c22996980206ce3849e1a3
Signed-off-by: C.J. Collier <cjcollier@linuxfoundation.org>
|
|
Change-Id: I7e792a561e8bf45f7b011b4dd85a69b96db4361d
Signed-off-by: Thanh Ha <thanh.ha@linuxfoundation.org>
|
|
Change-Id: Ia3a342e96349fc7c3cfd7448fce3e20e957bebff
Signed-off-by: Ed Warnicke <eaw@cisco.com>
|
|
Change-Id: I33ab63dc4870054253d8310c5fbc381e786e05f5
Signed-off-by: Ed Warnicke <eaw@cisco.com>
|
|
Change-Id: I13e35fcaf764db0c81dc335c0adb7e31ce7f8f42
Signed-off-by: Ed Warnicke <eaw@cisco.com>
|
|
Change-Id: Iebb012635d1eb2d428c6859d4cf6d7d962203995
Signed-off-by: Ed Warnicke <eaw@cisco.com>
|
|
Change-Id: I3c2a3650f30bd5d2bd49df30e78c57c5eb6d2ca8
Signed-off-by: Ed Warnicke <eaw@cisco.com>
|
|
Change-Id: I0d341f1cc96c3d8a35a127f7cd6f667126dac9d7
Signed-off-by: Ed Warnicke <eaw@cisco.com>
|
|
Change-Id: If0c4bfaf9aeb650b5fb13f4c1ac64bc4919b7c5f
Signed-off-by: Ed Warnicke <eaw@cisco.com>
|
|
Change-Id: Ibf5610b0607001178791bd9b56bcd0427e5ffee8
Signed-off-by: Ed Warnicke <eaw@cisco.com>
|
|
Change-Id: I10da36d45ff8ea3d4b5c3a17e68d5ee89d3132de
Signed-off-by: Ed Warnicke <eaw@cisco.com>
|
|
Change-Id: I1dd82bb2a8c5f7c19aa8905a24133e5805f31ae2
Signed-off-by: Ed Warnicke <eaw@cisco.com>
|
|
Change-Id: Ie1f34c60b256588281e70b835b7bbd797265db3a
Signed-off-by: Ed Warnicke <eaw@cisco.com>
|