apex.git
5 years agoRemoves the nova patch for NFS 21/67321/1 stable/gambia
Tim Rozet [Mon, 25 Feb 2019 20:33:52 +0000 (15:33 -0500)]
Removes the nova patch for NFS

Since we are not doing ODL csit anymore, can remove the complicated
patching for all nova services.

Change-Id: I44724a4bed7e595b03a0cddf14a7d4eb901e54ab
Signed-off-by: Tim Rozet <trozet@redhat.com>
(cherry picked from commit 7aac412d1bff29f1571cee6c38f1184ad127d1df)

5 years agoUpdates versioning for Gambia SR2 45/66645/1 opnfv-7.2.0
Tim Rozet [Fri, 25 Jan 2019 20:20:47 +0000 (15:20 -0500)]
Updates versioning for Gambia SR2

Includes updates to spec and docs.

Change-Id: Iba402177e4525b85d511a3d6406dd97b6a981897
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoAlways update OVS to 2.9.2 33/66033/1
Tim Rozet [Fri, 21 Dec 2018 17:36:36 +0000 (12:36 -0500)]
Always update OVS to 2.9.2

Right now RDO defaults to an old version of OVS. In Apex for non-sfc
scenarios we were pulling a prebuilt RPM of OVS 2.9.0 that included a
group tables fix needed for ODL. For SFC, we now just build OVS along
with the kernel module from native OVS because NSH support exists
upstream in that version.

We can move to always doing this operation for all OVS based scenarios,
which will bump our OVS and ensure the package we build at runtime will
work with our kernel.

Change-Id: If38d0b858878f1af66990d987612a8f8f6fa2718
Signed-off-by: Tim Rozet <trozet@redhat.com>
(cherry picked from commit 4f988328adf5dd8735bbf0a9420528cd343cfcd3)

5 years agoDisable NFS when all-in-one deployment 25/66025/1
Tim Rozet [Fri, 21 Dec 2018 14:34:05 +0000 (09:34 -0500)]
Disable NFS when all-in-one deployment

No point in using NFS when we only have one node.

Change-Id: Ic80f75913bf31c2e4a875146731d9c0d3483ae1b
Signed-off-by: Tim Rozet <trozet@redhat.com>
(cherry picked from commit 56a2abfb7a1b23e1716beb16702d130226e4778d)

5 years agoAttempting to fix NFS issues 91/65991/3
Tim Rozet [Sat, 15 Dec 2018 18:41:31 +0000 (13:41 -0500)]
Attempting to fix NFS issues

Issues still persist where sometimes instances fail to start due to a
failure with os.utime to read the file path. This could be some bad race
condition between qemu/nova while copying images on the NFS. This patch
adds more ports to open in firewall, and changes initial directory owner
to nfsnobody.

Also, includes a patch to fix an apparent race condition when nova sends
a remote call to the privsep helper daemon to modify the time of the
base file owned by qemu:
https://review.openstack.org/#/c/625741/

Includes another fix for patching container images where the docker
image was not being detected correctly because the full gerrit project
name including 'openstack/' prefix was being used to search tripleo
docker images. Additionally, there were more bugs around patching
openstack python containers where the patch was not being applied
correctly.

JIRA: APEX-654

Change-Id: I1d011035486298d5906038922e69d478c383c3f7
Signed-off-by: Tim Rozet <trozet@redhat.com>
(cherry picked from commit e1d286e89e04577bda2569a5909dfe8182d953ba)

5 years agoFix NFS issues with Nova 63/65863/1
Tim Rozet [Fri, 14 Dec 2018 17:59:09 +0000 (12:59 -0500)]
Fix NFS issues with Nova

There are problems with Nova launching instantces due to permissions
with nova being able to read/write certain directories on the NFS. The
permissions are right on the NFS and the folders the NFS mounts to, but
there still seems to be issues. The cause may be using a directory under
/root as the NFS mount. This patch moves the NFS mounts to be individual
folders under /. The patch also restart nova_compute docker container as
NFS problems still persist unless this is done.

JIRA: APEX-654

Change-Id: I25eee98c1a6516dfa44c686c2e614f6dc7000d98
Signed-off-by: Tim Rozet <trozet@redhat.com>
(cherry picked from commit 32b3b5dd6290ae5c33edee2860a3edd9f3044d43)

5 years agoBump RPM version for Gambia SR1 55/65855/1 opnfv-7.1.0
Tim Rozet [Fri, 14 Dec 2018 17:55:37 +0000 (12:55 -0500)]
Bump RPM version for Gambia SR1

Bumping version to 7.1 for Apex RPM.

Change-Id: Ide0333e15dfe73aac4954056f4586a9ddb840425
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoBump versions for 7.1 release 53/65853/1
Tim Rozet [Fri, 14 Dec 2018 17:50:44 +0000 (12:50 -0500)]
Bump versions for 7.1 release

Version bump for docs in prep for Gambia SR1.

Change-Id: Iba4abf1e3196072e4b0228f9874aa14c57adc09f
Signed-off-by: Tim Rozet <trozet@redhat.com>
(cherry picked from commit 654de7804b2264ef4729c22caa82f2d821938513)

5 years agoFixes mount issues with NFS 17/65717/1
Tim Rozet [Mon, 10 Dec 2018 16:31:33 +0000 (11:31 -0500)]
Fixes mount issues with NFS

There were issues related to moving to NFS for CSIT that included random
failures when mounting the NFS. Big thanks to radez for figuring out
that the issue was because the mounts were sharing the same root dir,
and thus the same context. Using nosharecache fixes the problem.

JIRA: APEX-654

Change-Id: Ic52d91a4abfe3db6289ce6a880c1f85ce5af7d0e
Signed-off-by: Tim Rozet <trozet@redhat.com>
(cherry picked from commit 0e0a291295998e77820e4b55f380d845fb509842)

5 years agoFix broken Gambia repos 13/65713/3
Tim Rozet [Mon, 10 Dec 2018 19:20:44 +0000 (14:20 -0500)]
Fix broken Gambia repos

The undercloud disk image repos look to be stale. This patch updates the
repos at deploy time to ensure the latest repos are always used. Similar
to what was done in:
https://gerrit.opnfv.org/gerrit/#/c/64493/

JIRA: APEX-656

Change-Id: I4eb2d0a8486b47a583434a9b83e1cc9d7274d9ad
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoMerge "Fix using ODL versions other than oxygen" into stable/gambia
Tim Rozet [Sat, 8 Dec 2018 18:35:19 +0000 (18:35 +0000)]
Merge "Fix using ODL versions other than oxygen" into stable/gambia

5 years agoFix using ODL versions other than oxygen 71/65671/1
Tim Rozet [Fri, 7 Dec 2018 14:49:21 +0000 (09:49 -0500)]
Fix using ODL versions other than oxygen

Fixes a bug where oxygen was always being used even if a different ODL
version was specified in the deploy settings. Also updates the valid ODL
versions with fluorine and neon.

JIRA: APEX-648

Change-Id: Ie49c28d1757cd7e57802d67e48904ce04864feaf
Signed-off-by: Tim Rozet <trozet@redhat.com>
(cherry picked from commit 46bf00b967844dc712dfe409e55fc5fe18fa119f)

5 years agoEnable CSIT deployments to use NFS backend 47/65647/2
Tim Rozet [Wed, 5 Dec 2018 14:35:52 +0000 (09:35 -0500)]
Enable CSIT deployments to use NFS backend

In order to support HA as well as live migration a shared storage
backend needs to be used. This patch preps the overcloud image to start
with an NFS server sharing out of /root/nfs. If csit environment file is
used then NFS parameters will be passed to the deployment, configuring
Nova and Glance to use NFS hosted by controller0.

JIRA: APEX-654

Change-Id: I049d84bfd454013fc7ba5048939ee17100197624
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoUpgrade OVS in undercloud 65/65665/4
Tim Rozet [Fri, 7 Dec 2018 21:09:50 +0000 (16:09 -0500)]
Upgrade OVS in undercloud

The OVS in the frozen undercloud disk image for Gambia is no longer
working (perhaps due to kernel upgrade or package changes during
deployment). This patch builds OVS the same way as is done for the SFC
scenario in the undercloud disk image.

JIRA: APEX-655

Change-Id: I6fa3e009a25806e299c7be88d6d3ee8db4c2648b
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoUse correct URL for opnfv-apex.repo 43/65543/1
Daniel Balsiger [Sun, 2 Dec 2018 15:40:05 +0000 (16:40 +0100)]
Use correct URL for opnfv-apex.repo

JIRA: APEX-652

Change-Id: I4651707265c1184026cf5ed06805997615d2db95
Signed-off-by: Daniel Balsiger <Daniel.Balsiger@swisscom.com>
(cherry picked from commit 5c41b8486a46e3c86a573b7130792624b3f5ed26)

5 years agoMake util support more than 10 compute nodes 33/65233/1
Daniel Balsiger [Fri, 2 Nov 2018 22:00:46 +0000 (23:00 +0100)]
Make util support more than 10 compute nodes

JIRA: APEX-644

The grep command used in the original code matches multiple nodes
when deploying 10+ compute nodes. This makes opnfv-util fail to
connect to those nodes. Proposed fix is to grep for "$node "
instead.

Change-Id: I952664f6f3cab3a98681684cd354513bb3c0ab12
Signed-off-by: Daniel Balsiger <Daniel.Balsiger@swisscom.com>
(cherry picked from commit 859fb4703b3f0359e80b35aaa62a94e752707bae)

5 years agoUpdates documentation for Gambia 03/64903/1 opnfv-7.0.0
Tim Rozet [Fri, 9 Nov 2018 16:17:28 +0000 (11:17 -0500)]
Updates documentation for Gambia

Changes-Include:
 - Remove references to Apex ISO and disk image RPMs
 - Update supported scenarios
 - Update using upstream documentation

Change-Id: If2b66d1d5a861bd1f90b0e8829de37d84e656619
Signed-off-by: Tim Rozet <trozet@redhat.com>
(cherry picked from commit b188263699077102e956929970cc20191dafd312)

5 years agoFix Undercloud URL path 03/64603/1
Tim Rozet [Tue, 6 Nov 2018 14:56:21 +0000 (09:56 -0500)]
Fix Undercloud URL path

Modify the URL path for undercloud disk image to opnfv artifacts. The
undercloud image was removed from upstream TripleO, so we need to use
the latest known good version for Gambia release, stored in OPNFV
artifacts. In the future we will remove downloading undercloud image and
just use the overcloud image for undercloud.

Change-Id: I17a3bbaf67d55de05c195712d7fee6e89676bc16
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoAdds SDN Port variables to overcloudrc 57/64357/1
Tim Rozet [Tue, 23 Oct 2018 14:50:25 +0000 (10:50 -0400)]
Adds SDN Port variables to overcloudrc

Functest needs these values to be set in overcloudrc to know which ports
to query ODL on.

JIRA: APEX-621

Change-Id: I41e34efccedc26edd98c6dd3f45e553ea76db195
Signed-off-by: Tim Rozet <trozet@redhat.com>
(cherry picked from commit 63792906e8681bef2f0e03ee12ea422af53d9c12)

5 years agoChange OPNFV Artifacts URL 31/64031/2
Tim Rozet [Wed, 24 Oct 2018 16:25:20 +0000 (12:25 -0400)]
Change OPNFV Artifacts URL

This change allows downloading of files larger than 20GB.

Change-Id: I881ea35316e67980d4144e46bfbd41501e80f497
Signed-off-by: Tim Rozet <trozet@redhat.com>
(cherry picked from commit 498356c9efb765a5714845ce8303dc8a264eb300)

5 years agoDowngrade ceph-ansible package 07/64307/1
Tim Rozet [Wed, 31 Oct 2018 13:53:59 +0000 (09:53 -0400)]
Downgrade ceph-ansible package

Ceph Ansible is failing deployments with:
https://bugzilla.redhat.com/show_bug.cgi?id=1644713

Change-Id: I4e1cad9330f77472e52b5d790b008e9d238cc48e
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoGambia Branch Configuration for Docs and Gitreview 17/63917/1
Trevor Bramwell [Tue, 23 Oct 2018 16:25:02 +0000 (09:25 -0700)]
Gambia Branch Configuration for Docs and Gitreview

Update the docs to use 'Gambia' as the version and release, and sets the
default gitreview branch to 'stable/gambia' so it doesn't need to be
explicly passed every time.

Change-Id: Id5cbfb581cb232af5541f58df8c95e4b0918713e
Signed-off-by: Trevor Bramwell <tbramwell@linuxfoundation.org>
5 years agoFixes undercloud accidentally using wrong images/containers 63/63663/4
Tim Rozet [Wed, 17 Oct 2018 15:17:41 +0000 (11:17 -0400)]
Fixes undercloud accidentally using wrong images/containers

Our undercloud initial install was failing with queens because the
--use-heat arg does not work the same way as master/rocky. Therefore the
undercloud install would fail, and then try to reinstall on the next
ansible task. This one would succeed, via container installation because
we were using the wrong undercloud/overcloud disk images.

Change-Id: Ifb829aeeec991ba6dbe582fda3e9ee6b9bc733df
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoLocal Documentation Builds 39/63139/1
Aric Gardner [Tue, 18 Sep 2018 16:10:43 +0000 (12:10 -0400)]
Local Documentation Builds

This adds configuration for performing local documentation builds
with the following simple command:

tox -e docs

This patch is part of our documentation effor to Move away from submodules
https://docs.opnfv.org/en/latest/how-to-use-docs/local-build-transition.html

added index.rst so that docs build succsesfully

Change-Id: I2d28f03ad4d5c9c14de095dc544257af9d1e51d9
Signed-off-by: Aric Gardner <agardner@linuxfoundation.org>
(cherry picked from commit 9666d8bb3df726d88d08d5811d2a99e713908396)

5 years agoMigrates deploy settings for Gambia 37/63037/2
Tim Rozet [Fri, 28 Sep 2018 19:42:49 +0000 (15:42 -0400)]
Migrates deploy settings for Gambia

With branching to gambia we need to move the settings back to their
proper names and remove OpenStack branch as Queens is the default for
Gambia.

Change-Id: Ia6101289629d89fa349f6554436ccd7c0f8b3c0c
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoEnable OVN scenarios 33/63033/1
Tim Rozet [Tue, 21 Aug 2018 13:19:57 +0000 (09:19 -0400)]
Enable OVN scenarios

As of Queens only HA OVN deployments are supported.

Change-Id: I184c5a096fec9cbc3cf2ec06218700138ea3ed57
Signed-off-by: Tim Rozet <trozet@redhat.com>
(cherry picked from commit ee58a24647df8eec2418735b038e5035bdffe10d)

5 years agoFix per-network routes to NIC templates dependency 13/62113/8
Ricardo Noriega [Tue, 11 Sep 2018 16:28:29 +0000 (18:28 +0200)]
Fix per-network routes to NIC templates dependency

Change-Id: I9e01f1164fc72915b92dfb1c0aad7414c484567e
Signed-off-by: Ricardo Noriega <rnoriega@redhat.com>
5 years agoUpdates Calipso deploy settings 47/61347/8
Tim Rozet [Fri, 24 Aug 2018 16:46:55 +0000 (12:46 -0400)]
Updates Calipso deploy settings

Change-Id: Ibfbd08dc2fa5fca95668fd0590707cfebd92099f
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoAdds deployment via snapshot 43/60543/15
Tim Rozet [Fri, 3 Aug 2018 03:49:00 +0000 (23:49 -0400)]
Adds deployment via snapshot

New arguments are added to allow snapshot deployment:
--snapshot, --snap-cache

The previous tripleo-quickstart code has been removed/replaced
with the snapshot option.

Snapshot deployments are supported on CentOS and Fedora, and snapshot
artifacts use a similar caching system as the standard deployment.
Snapshots are produced daily by Apex, and include latest as well as n-1
OpenStack versions. The os-odl-nofeature scenario is used for the
snapshots. Additionally multiple topology verions of Snapshots are
available. The Snapshot pulled at deploy time depends on the
deploy-settings and number of virtual-computes used at deploy time.

Since there is only one network used with snapshot deployments (admin),
there is no reason to pass in network settings for snapshot deployments.
That argument is now optional. Previously we required even in Standard
virtual deployments that the network settings be provided. However that
is also unnecessary, as we can default to the virtual network settings.

Includes minor fix to the tox.ini to allow specifying test cases
to run (useful for developers writing tests). Default behavior of
tox is unchanged.

JIRA: APEX-548

Change-Id: I1e08c4e54eac5aae99921f61ab7f69693ed12b47
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoRecover fix to install custom OVS 01/61301/3
Ricardo Noriega [Thu, 23 Aug 2018 17:10:01 +0000 (19:10 +0200)]
Recover fix to install custom OVS

  Need this fix to make VLAN tenant networks
  to work.

Change-Id: Ie2dea9f0145c576e1b36280e6340ed2cc0460179
Signed-off-by: Ricardo Noriega <rnoriega@redhat.com>
5 years agoMerge "Enable SFC scenarios for Gambia"
Tim Rozet [Wed, 22 Aug 2018 21:31:53 +0000 (21:31 +0000)]
Merge "Enable SFC scenarios for Gambia"

5 years agoEnable SFC scenarios for Gambia 15/60715/17
Ricardo Noriega [Tue, 7 Aug 2018 16:29:54 +0000 (18:29 +0200)]
Enable SFC scenarios for Gambia

  - This patch will install OVS 2.9.2 including
    its kernel module which allows native NSH
    headers.
  - Fix Custom OVS due to bug:
    https://bugzilla.redhat.com/show_bug.cgi?id=1544892
  - Tacker is disable for the time being, tacker-conductor
    needs to be enabled.

JIRA: APEX-630

Change-Id: Ia410309fd7053602ce78eae919839d0f57c9742a
Signed-off-by: Ricardo Noriega <rnoriega@redhat.com>
5 years agoFixes unit test failures 25/61125/1
Tim Rozet [Mon, 20 Aug 2018 16:34:43 +0000 (12:34 -0400)]
Fixes unit test failures

Adds mocking some file opens introduced by:
https://gerrit.opnfv.org/gerrit/#/c/60991/

Change-Id: I013b9c96a7c3f678c6e227f98cd58f9547d79ecb
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoMerge "Add Nova migration params to upstream env file"
Tim Rozet [Mon, 20 Aug 2018 13:12:05 +0000 (13:12 +0000)]
Merge "Add Nova migration params to upstream env file"

5 years agoAdd Nova migration params to upstream env file 91/60991/5
Ricardo Noriega [Thu, 16 Aug 2018 07:39:01 +0000 (09:39 +0200)]
Add Nova migration params to upstream env file

JIRA: APEX-635

Change-Id: I488dd5d9b114aba5f4bb4591a0cdd8c19b692ce9
Signed-off-by: Ricardo Noriega <rnoriega@redhat.com>
5 years agoFix ipv6 deployment failures 09/61009/3
Feng Pan [Thu, 16 Aug 2018 15:36:58 +0000 (11:36 -0400)]
Fix ipv6 deployment failures

- fix ipv6 enable command on bridge interfaces
- fix external subnet creation command
- enable NAT when admin is ipv4
- fix network settings file for ipv6 for br-ex interface type

Change-Id: Ibd2dbe1d0c8e8df72ad1b0b3713a19f8bbc56eef
Signed-off-by: Feng Pan <fpan@redhat.com>
5 years agoModify common patches schema to per branch 65/60965/3
Tim Rozet [Wed, 15 Aug 2018 01:17:33 +0000 (21:17 -0400)]
Modify common patches schema to per branch

The common patch design doesn't really work across branches very well.
This patch makes it so the common patches are specified in the same
file, but on a per branch basis. Making it easier to manage common
patches per branch.

Also, includes a fix and adds test coverage to the prep_image method
where we were referencing 'undercloud_admin_ip' before assignment in a
case where we were not using ODL and had patches to overcloud.

Change-Id: I7672947afd826fdc0042361a0139e22d6d5dd864
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoAllow common patches file 09/60909/7
Tim Rozet [Mon, 13 Aug 2018 18:51:09 +0000 (14:51 -0400)]
Allow common patches file

This patch adds allowing for common patches that should be applied to
every scenario to be included. It by default pulls in a file in the
deploy directory 'common-patches.yaml', but can optionally be
overridden.

This patch also includes a patch upstream to fix OSCLI not working
anymore due to breakage with the Cinder version in the overcloudrc.

Change-Id: I97b9efb937deff07e085b9ef75b9799fb65bfc57
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoAllow all in one deployments 51/60751/5
Tim Rozet [Wed, 8 Aug 2018 21:43:55 +0000 (17:43 -0400)]
Allow all in one deployments

This patch adds the ability to deploy all in one single nodes (Control
+ Compute). To enable this functionality do the following for each
deployment type:
 - Baremetal: do not tag any nodes as compute in the inventory file
 - Virtual: use argument '--virtual-computes 0'

JIRA: APEX-548

Change-Id: I22525c9eb21d331129c819449316c26a6fcf522d
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoFix broken clean due to virtualbmc change 07/60807/1
Tim Rozet [Thu, 9 Aug 2018 20:38:08 +0000 (16:38 -0400)]
Fix broken clean due to virtualbmc change

Apex clean is failing when trying to clean vbmcs because the function
was changed in new virtualbmc version. Adding the working version to our
constraints.

JIRA: APEX-631

Change-Id: I62971183669e4a51e841d79892e78b10ea380a1e
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoBump deploy timeout 11/60711/4
Tim Rozet [Tue, 7 Aug 2018 15:50:37 +0000 (11:50 -0400)]
Bump deploy timeout

We are seeing queens baremetal deployments are all failing in step 5 due
to timeout. Not sure why the sudden increase in deployment time, but we
can try increasing the deploy timeout to see if it fixes it.

Also includes a revert for a commit that broke OOO master:
https://bugs.launchpad.net/tripleo/+bug/1785872

Change-Id: I0bcfd5b3ca49b176f901c13b8b432fa29c899188
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoRevert "Fix BGPVPN Queens scenario files to be consistent" 15/60615/1
Ricardo Noriega [Mon, 6 Aug 2018 13:45:32 +0000 (13:45 +0000)]
Revert "Fix BGPVPN Queens scenario files to be consistent"

This reverts commit b88eb3c7a1578cb251282c358fad6d87e7439868.

Change-Id: If5a2548e647de099229824761c6c575ca4f2e815
Signed-off-by: Ricardo Noriega <rnoriega@redhat.com>
5 years agoFix BGPVPN Queens scenario files to be consistent 67/60567/1
Ricardo Noriega [Fri, 3 Aug 2018 13:18:06 +0000 (15:18 +0200)]
Fix BGPVPN Queens scenario files to be consistent

Change-Id: Id5b932246020676e1e53438db689e32bc124939e
Signed-off-by: Ricardo Noriega <rnoriega@redhat.com>
5 years agoEnable BGPVPN for master deployments 69/60269/18
Ricardo Noriega [Fri, 27 Jul 2018 09:29:41 +0000 (11:29 +0200)]
Enable BGPVPN for master deployments

  - Injection of Quagga tarball via overcloud builder.
  - Extraction and installation of all related packages.
  - It uses SDNVPN artifact repository to download Quagga
    tarball, so there is only one source to test.
  - Modifies bgpvpn scenario files to use OS master branch,
    ODL master branch and containers.

JIRA: APEX-627

Change-Id: Icdbc2853d9531048e23fd6d5e444bd68208d18fc
Signed-off-by: Ricardo Noriega <rnoriega@redhat.com>
5 years agoFixes missing NovaPlacement for csit 79/60479/1
Tim Rozet [Wed, 1 Aug 2018 13:10:36 +0000 (09:10 -0400)]
Fixes missing NovaPlacement for csit

Nova Placement should have been on controller.

Change-Id: I9279e56b5cf786c1153af4b09b29e2b215c2f248
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoUse metadata IP instead of FQDN 47/60447/1
Tim Rozet [Tue, 31 Jul 2018 20:14:25 +0000 (16:14 -0400)]
Use metadata IP instead of FQDN

There is an issue with loss of external network connectivity that
prevents cloud init to instances working. This becomes a big problem
with snapshots where there is no external network connectivity. Cloud
init fails because each request takes over 30 seconds to get a response.
This is because in the background neutron metadata agent is proxying the
request to nova metadata agent with an HTTP GET using the FQDN. For
whatever reason, a DNS lookup happens even though the entry exists for
the FQDN in /etc/hosts and waits 30 seconds until timing out. After this
timeout, a 200 OK is sent and metadata works.

This patch modifies the config file for metadata to use nova metadata
server's internal IP rather than FQDN as there is no option in OOO to
use IPs instead of FQDN.

Change-Id: I6960181a227d0002c99aeae5112f59807dc41d7a
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoUpdates CSIT env files 91/60291/1
Tim Rozet [Fri, 27 Jul 2018 16:38:47 +0000 (12:38 -0400)]
Updates CSIT env files

NovaPlacementApi is not needed on compute nodes and is not default on
compute nodes upstream, so removing it. Also adds clustercheck and sql
clients to the both roles.

Change-Id: Ia9d9fb4f115bc9850be46043a03688b55149f5df
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoMerge "Increasing loop device size for baremetal deployments"
Tim Rozet [Fri, 27 Jul 2018 02:01:50 +0000 (02:01 +0000)]
Merge "Increasing loop device size for baremetal deployments"

5 years agoIncreasing loop device size for baremetal deployments 07/60207/6
Ricardo Noriega [Thu, 26 Jul 2018 09:17:37 +0000 (11:17 +0200)]
Increasing loop device size for baremetal deployments

Change-Id: Ib3779733ac27183def6a59f3404cfbe9a3e9a211
Signed-off-by: Ricardo Noriega <rnoriega@redhat.com>
5 years agoUpdate OVS in overcloud 17/60217/3
Tim Rozet [Thu, 26 Jul 2018 18:46:24 +0000 (14:46 -0400)]
Update OVS in overcloud

This bug fix is not present in the RDO versions of OVS:
https://bugzilla.redhat.com/show_bug.cgi?id=1544892

The impact of this is that traffic sent through a patch port (like from
br-int to br-ex) will not be sent correctly with OF groups (which ODL
uses). This would cause tenant vlan traffic not to work in our case.

Change-Id: Ia477b174118dcd2d7fb24189c2d093218eb97053
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoDisable undercloud as containers for now 19/60219/2
Tim Rozet [Thu, 26 Jul 2018 19:31:20 +0000 (15:31 -0400)]
Disable undercloud as containers for now

Upstream has moved to install undercloud as containers which breaks our
post undercloud configuration for some services. Disable it for now and
then move to it gracefully in a future patch.

Change-Id: I0cd1a8ddac4ba92734750265d2a16d3ef008f236
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoMerge "Remove obsolete Ceph tags"
Tim Rozet [Thu, 26 Jul 2018 19:01:30 +0000 (19:01 +0000)]
Merge "Remove obsolete Ceph tags"

5 years agoRemove obsolete Ceph tags 09/60109/9
Ricardo Noriega [Mon, 23 Jul 2018 10:14:59 +0000 (12:14 +0200)]
Remove obsolete Ceph tags

  This patch removes the logic to use an specific tag for
  Ceph containers. We will use whatever docker image TripleO
  upstream uses. For aarch64, an ansible task will replace
  the tag to pull the proper container image.

  This patch also refactors the preparation of the local
  registry. In Queens, there is no need to execute twice
  the overcloud container image prepare command.

JIRA: APEX-622

Change-Id: I947d931609e58505675bb460a59d08c1d10d1d0b
Signed-off-by: Ricardo Noriega <rnoriega@redhat.com>
5 years agoOpen port 8101 on controllers for karaf 63/60163/3
Tim Rozet [Tue, 24 Jul 2018 15:21:06 +0000 (11:21 -0400)]
Open port 8101 on controllers for karaf

By default 8101 (karaf shell) is blocked on controllers. In Apex we
advertise in our user guide (and tools scripts) the ability to connect
to karaf shell. It is also required to run CSIT. This patch opens the
port when ODL is deployed.

Change-Id: Ib3ece41f19607bafc329d9de390cf774766a26cd
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoMerge "Fixes for snapshots"
Tim Rozet [Sat, 21 Jul 2018 03:34:09 +0000 (03:34 +0000)]
Merge "Fixes for snapshots"

5 years agoFixes for snapshots 53/60053/3
Tim Rozet [Fri, 20 Jul 2018 14:34:39 +0000 (10:34 -0400)]
Fixes for snapshots

With deploying snapshots with a new ODL, we currently bring down the
docker container and bring up the tar.gz distro of ODL on the Overcloud
host itself (not rebuilding/using container). Therefore we need java
installed so that ODL can run on the host. In the future this may
change, but it works well and keeps things simple for now.

Additionally, there was a change upstream to make the opendaylight
container docker restart policy "unless-stopped" which means it will
no longer restart automatically when docker is stopped/started.
Therefore on first snapshot bring up (without the previously mentioned
ODL reinstallation) the container does not start, and snapshot
deployment fails. This patch includes a change to the restart policy to
always restart it.

Change-Id: Icc712ba147e578a28e371313154ae3190676f0dc
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoFixes missing csit queens file in setup.cfg 51/60051/1
Tim Rozet [Fri, 20 Jul 2018 13:24:23 +0000 (09:24 -0400)]
Fixes missing csit queens file in setup.cfg

Change-Id: I6d57c88efc88dba311e79a36024683d5a9ade679
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoMerge "Add tenant vlan support"
Tim Rozet [Fri, 20 Jul 2018 00:24:47 +0000 (00:24 +0000)]
Merge "Add tenant vlan support"

5 years agoAdd tenant vlan support 87/59287/6
Feng Pan [Fri, 29 Jun 2018 03:48:05 +0000 (23:48 -0400)]
Add tenant vlan support

Introduces tenant vlan support through configuration of network settings file.

JIRA: APEX-272

Change-Id: Ie8e20a92ca4ebb9e0de4c53b5e03856d19ff660c
Signed-off-by: Feng Pan <fpan@redhat.com>
5 years agoAdd pacemaker for HA snapshot deployments 59/59959/1
Tim Rozet [Wed, 18 Jul 2018 18:39:08 +0000 (14:39 -0400)]
Add pacemaker for HA snapshot deployments

Change-Id: Ic7b631af3968de8868177efa719f7c6e01628ce9
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoAdd param for ODL password into overcloudrc 11/59811/3
Tim Rozet [Fri, 13 Jul 2018 20:22:30 +0000 (16:22 -0400)]
Add param for ODL password into overcloudrc

Recent changes upstream have removed the default 'admin' ODL password
and now password is randomly generated:

https://review.openstack.org/#/c/578505/

So in OPNFV we now store the password in overcloudrc as
SDN_CONTROLLER_PASSWORD variable.

Also includes minor fixes to unittests.

Change-Id: Iabe7e4f902442c80af99ba1603a3927cf13d0393
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoUpdate queens ODL to include OF table fix 69/59869/3
Tim Rozet [Mon, 16 Jul 2018 17:49:21 +0000 (13:49 -0400)]
Update queens ODL to include OF table fix

Also includes minor fix to patching where stable was not being removed
from the short hand branch name correctly.

Includes fix where path was changed in queens for env files:
https://review.openstack.org/#/c/567810/

Change-Id: Ia3a6769be70ea97c9bec9d5539c9ab0373fed30a
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoMerge "Add support for kubernetes deployment"
Tim Rozet [Mon, 16 Jul 2018 14:14:26 +0000 (14:14 +0000)]
Merge "Add support for kubernetes deployment"

5 years agoFix OpenDaylight Oxygen deployment failures (master) 97/59797/1
Tim Rozet [Fri, 13 Jul 2018 15:25:20 +0000 (11:25 -0400)]
Fix OpenDaylight Oxygen deployment failures (master)

ODL scenarios are failing deployment with oxygen because table 17 from
the OpenFlow pipeline is missing. This pulls in an upstream patch to no
longer check for table 17. The default table 17 flow has been removed
upstream which is why it no longer appears during deployment.

This patch fixes master OS branch scenarios only. Once the patch is
cherry-picked to queens we can submit a new patch updating the DS
settings.

Change-Id: If60cf25c75440e3a1998b412ac8b903849507dab
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoMerge "Removes OC condition in dhcp-all-interfaces"
Tim Rozet [Thu, 12 Jul 2018 16:37:35 +0000 (16:37 +0000)]
Merge "Removes OC condition in dhcp-all-interfaces"

5 years agoAdds IP to external network NIC on computes 51/59651/1
Tim Rozet [Tue, 10 Jul 2018 14:10:56 +0000 (10:10 -0400)]
Adds IP to external network NIC on computes

Deployments now fail in baremetal because the external NTP server cannot
be reached because there is no IP on the extenal interfaces for
computes.

Change-Id: Ic1e67072b81bb7ca4e15930840dc77d27913737f
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoAdd support for kubernetes deployment 77/53077/27
Zenghui Shi [Wed, 4 Apr 2018 03:24:40 +0000 (11:24 +0800)]
Add support for kubernetes deployment

This patch adds capability to deploy kubernetes cluster instead of openstack.
Kubernetes will be deployed using kubespray and is run after TripleO bookstraps
overcloud nodes.

JIRA: APEX-574

Change-Id: If9c171620c933a052b719e7112a50e22bbab667f
Signed-off-by: Feng Pan <fpan@redhat.com>
Signed-off-by: Zenghui Shi <zshi@redhat.com>
5 years agoPhysical interface type note 57/59557/2
Tomi Juvonen [Thu, 5 Jul 2018 06:07:41 +0000 (09:07 +0300)]
Physical interface type note

add ovs_bridge as needed compared to earlier version

Change-Id: I1d0c6df9d1f47b0c98a3c1a8bf3a04eb5f53aff2
Signed-off-by: Tomi Juvonen <tomi.juvonen@nokia.com>
5 years agoFix neutron-opendaylight-sriov.yaml path 83/59583/1
Feng Pan [Fri, 6 Jul 2018 14:07:14 +0000 (10:07 -0400)]
Fix neutron-opendaylight-sriov.yaml path

In latest THT upstream, environment file neutron-opendaylight-sriov.yaml was
moved to services folder. Updating references in Apex to avoid deploy failure.

Change-Id: I7065e0d8e13c9add9ead282db2244a27c177e5a4
Signed-off-by: Feng Pan <fpan@redhat.com>
5 years agoDeprecates the use of '--upstream' argument 11/59011/2
Tim Rozet [Mon, 25 Jun 2018 19:21:07 +0000 (15:21 -0400)]
Deprecates the use of '--upstream' argument

We only support the use of upstream deployments from Gambia onwards.
Therefore deprecating the parameter and forcing upstream to always be
used.

Change-Id: I7926a65a2c32bca2272855ba45fdb6005f67616f
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoIgnore error if network already undefined 13/59013/1
Tim Rozet [Mon, 25 Jun 2018 20:05:33 +0000 (16:05 -0400)]
Ignore error if network already undefined

In clean we destroy a network, then undefine it. When running snapshot
deploy it uses manual virsh bash commands to create the networks and
other virsh resources. This leaves an admin network which our clean
eventually tries to delete. However when the libvirt api network destroy
command is used it is somehow undefining the newtork at the same time.
Therefore just catch that possible scenario and ignore the failure as
the network has already been removed.

Change-Id: Id1d861fbe2a338f0828e78721c09f110030d3a4a
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoAuto Generated INFO.yaml file 21/51521/5
agardner [Fri, 2 Feb 2018 22:12:24 +0000 (17:12 -0500)]
Auto Generated INFO.yaml file

INFO.yaml meant to be machine readable.
This file was auto generated from an ldap lookup
and your projects INFO file

Meeting times are not currently filled out.
please ammend this patchset to complete

Going forward comitter promotions should be done
against this file.

PTLs:
Must be merged for you to gain delete access on
testresults.opnfv.org

Change-Id: Ia3749f55a407fa1dd173021b509123029bcde297
Signed-off-by: agardner <agardner@linuxfoundation.org>
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoMerge "Fixes Ceph PG calculation"
Feng Pan [Tue, 19 Jun 2018 14:47:26 +0000 (14:47 +0000)]
Merge "Fixes Ceph PG calculation"

5 years agoFixes Ceph PG calculation 83/58683/3
Tim Rozet [Mon, 18 Jun 2018 18:00:38 +0000 (14:00 -0400)]
Fixes Ceph PG calculation

Baremetal deployments were failing because the ceph PG size was
exceeding the max allowed. Virtual was still working because we lower
the number of pools and pg/osd. This patch changes the values to a
number which should work for both virtual and baremetal. Also includes a
fix which adds the controllers back as OSDs and a few other cleanup
issues.

JIRA: APEX-614
JIRA: APEX-569

Change-Id: I2ad65727ecdcaa0454eb53d25e32b7f1a53cd3a4
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoFetch mistral logs from undercloud 71/58671/2
Tim Rozet [Mon, 18 Jun 2018 15:28:35 +0000 (11:28 -0400)]
Fetch mistral logs from undercloud

/var/lib/mistral path contains logs for when ansible is invoked by
TripleO for Ceph configuration as well as config download. This patch
now archives and fetches that directory. Logs in previous releases
like Queens store the Ceph logs in /var/log/mistral.

Change-Id: I50c43e55efaa5dbcf8b7fb00b0e11cd3288fdd05
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoAdds missing queens deploy settings files 03/58603/1
Tim Rozet [Sat, 16 Jun 2018 02:54:52 +0000 (22:54 -0400)]
Adds missing queens deploy settings files

Change-Id: I87508e6d486ade6e4acd8dec69a4c867b0d81252
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoFixes deployment failure with allNodesConfig 17/58217/2
Tim Rozet [Tue, 5 Jun 2018 01:19:22 +0000 (21:19 -0400)]
Fixes deployment failure with allNodesConfig

This pulls in upstream patch to revert a bad commit that causes the
"Unknown Property controller_ips". Also includes a fix for being able to
detect if patches that are merged upstream are also promoted into
TripleO images or container images. This happens by comparing the time
the patch was submitted to the time when the TripleO Image or Docker
Image was last updated.

JIRA: APEX-610
JIRA: APEX-612

Change-Id: I1c2ab7fb4425b407acd7b6d9ebab914ed3a24478
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoMerge "Configure NAT with baremetal when necessary"
Feng Pan [Thu, 31 May 2018 17:32:27 +0000 (17:32 +0000)]
Merge "Configure NAT with baremetal when necessary"

5 years agoMerge "Updates configs for ODL CSIT deployments"
Feng Pan [Wed, 30 May 2018 19:56:33 +0000 (19:56 +0000)]
Merge "Updates configs for ODL CSIT deployments"

5 years agoConfigure NAT with baremetal when necessary 91/57791/3
Tim Rozet [Wed, 23 May 2018 20:51:48 +0000 (16:51 -0400)]
Configure NAT with baremetal when necessary

We currently only enable NAT on undercloud for virtual deployments.
However, there could be a case where a baremetal deployment also needs
NAT as it is not using an interface on the overcloud nodes with external
access. Therefore this patch changes the behavior to configure NAT when
the gateway of either the external or admin (when external is disabled)
network matches an IP assigned to the undercloud.

JIRA: APEX-605

Change-Id: I9c79af371913e6e5f0d39b433f68205bc7e106c5
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoUpdates configs for ODL CSIT deployments 75/54575/6
Tim Rozet [Wed, 28 Mar 2018 20:58:15 +0000 (13:58 -0700)]
Updates configs for ODL CSIT deployments

For ODL CSIT we want to deploy with:
 - Minimum services per role
 - 2 Compute nodes, 1 controller
 - Single network enabled

Change-Id: If611c4c1ff68629670ef15904930124b5786a569
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoMerge "Default ctlplane to native VLAN"
Feng Pan [Tue, 29 May 2018 17:13:01 +0000 (17:13 +0000)]
Merge "Default ctlplane to native VLAN"

5 years agoMerge "Remove pacmeaker workaround"
Feng Pan [Tue, 29 May 2018 13:10:59 +0000 (13:10 +0000)]
Merge "Remove pacmeaker workaround"

5 years agoMerge "Promote Ricky to committer"
Tim Rozet [Tue, 29 May 2018 13:09:24 +0000 (13:09 +0000)]
Merge "Promote Ricky to committer"

5 years agoRemove pacmeaker workaround 39/57939/1
Tim Rozet [Fri, 25 May 2018 19:54:08 +0000 (15:54 -0400)]
Remove pacmeaker workaround

There was a compatibility issue with the centos 7.4/7.5 between the host
pacemaker version and container.  Now that containers have moved to 7.5
we should not need this workaround anymore.

Change-Id: I9632c65e87687d4f36130719c6df9af2e913eed8
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoAdding git as required dependency for Apex RPMs 63/57463/2
Ricardo Noriega [Wed, 16 May 2018 10:54:40 +0000 (12:54 +0200)]
Adding git as required dependency for Apex RPMs

  If it's not installed, opnfv-deploy will output an
  error message.

JIRA: APEX-599

Change-Id: Ib826204120f53abce1f4f1e3e4ec3119a71ab650
Signed-off-by: Ricardo Noriega <rnoriega@redhat.com>
5 years agoDefault ctlplane to native VLAN 87/57787/1
Tim Rozet [Wed, 23 May 2018 20:09:47 +0000 (16:09 -0400)]
Default ctlplane to native VLAN

We were not setting any vlan type value for the admin network.  This
caused an issue when deploying and trying to use a single collapsed
network (no net isolation).  The issue occurred when trying to create an
external neutron network.  We happen to check if the phys type is flat
or vlan using this attribute to decide what kind of phys type to use for
the neutron network.

JIRA: APEX-606

Change-Id: I4e24dd5e8b99cef920b8203b820a77d0021631cc
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoPromote Ricky to committer 83/57783/1
Tim Rozet [Wed, 23 May 2018 19:42:22 +0000 (15:42 -0400)]
Promote Ricky to committer

Ricky has been a very active Apex user and contributor over the past few
release cycles.  He has helped find and fix numerous bugs, along with
being responsible for L2GW and SRIOV features for Fraser release.  He
has shown a willingness to jump in and help whenever there are issues
with Apex, and also a willingness to work upstream in TripleO/OpenStack
to implement new features.

Change-Id: Ifd032fd349a46b0e34b2d77ba56e3c564e70226c
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoFix RPM build 85/57685/2
Tim Rozet [Tue, 22 May 2018 13:56:50 +0000 (09:56 -0400)]
Fix RPM build

networking-vpp was referencing variables from the old variables.sh.
Moving those values into the Makefile as the variables.sh file was
removed.

Change-Id: I8ef5e6988299e7e3855d442657db2ed20086689f
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoFixes ctlplane default route 15/57415/2
Tim Rozet [Tue, 15 May 2018 20:38:44 +0000 (16:38 -0400)]
Fixes ctlplane default route

Default route was using the undercloud IP, and instead should be using
the gateway set for the network in network settings.

JIRA: APEX-597

Change-Id: Iff6b18a6553af98cf9da72c278f358922d489958
Signed-off-by: Tim Rozet <trozet@redhat.com>
5 years agoMigrates master to use direct upstream 81/56981/34
Tim Rozet [Wed, 2 May 2018 16:34:38 +0000 (12:34 -0400)]
Migrates master to use direct upstream

We now move master to deploy from upstream.  That means we do not need
to build undercloud/overcloud images anymore.

Changes-Include:
 - Remove bash build scripts as we do not need to build anything
   other than the python package anymore
 - Remove building images or iso from build.py
 - Remove building of images and iso from Makefile
 - Rename/refactor deploy settings files for nosdn and odl.  The new
   convention is that the typical scenario names we use will deploy
   master.  We also support n-1 OS, so in that case we use the branch
   name for the "feature" in the scenario name: os-odl-queens-noha.
 - Tacker/Congress are disabled in settings files until we fix that with
   upstream.  Containers are now enabled by default.
 - Disable TLS for undercloud (was changed upstream to default enabled)
 - Fix environments docker directory for master THT (was changed upstream)
 - Includes fix for LP#1768901
 - Includes workaround for LP#1770692
 - Moves to docker.io for container images as it is more stable and
   should contain the same images
 - Removes the term 'common' from apex packaging for referencing the
   Python Apex package

Change-Id: If6b433860b3ff882686c78d0f24a2f0c52b9b57a
Signed-off-by: Tim Rozet <trozet@redhat.com>
6 years agoRemoves OC condition in dhcp-all-interfaces 37/56937/1
Tim Rozet [Tue, 1 May 2018 20:46:58 +0000 (16:46 -0400)]
Removes OC condition in dhcp-all-interfaces

Due to this condition, a non-first ctlplane NIC may fail to come up in
deployments, causing deployment to fail.  This is a temporary workaround
until the proper fix can be merged:

https://review.openstack.org/#/c/561003/1

JIRA: APEX-588

Change-Id: I3f19afba2618161843ce3b76a03e2dcf1fcb4b99
Signed-off-by: Tim Rozet <trozet@redhat.com>
6 years agoUpdates documentation for Fraser release 23/56523/4
Tim Rozet [Wed, 25 Apr 2018 17:45:31 +0000 (13:45 -0400)]
Updates documentation for Fraser release

Change-Id: I19bfaa5b2778baf458df60a4a53d135e57859e04
Signed-off-by: Tim Rozet <trozet@redhat.com>
6 years agoMerge "Fix nosdn fdio scenario"
Tim Rozet [Thu, 19 Apr 2018 04:12:15 +0000 (04:12 +0000)]
Merge "Fix nosdn fdio scenario"

6 years agoMerge "Fixes OVN scenario"
Feng Pan [Thu, 19 Apr 2018 02:12:50 +0000 (02:12 +0000)]
Merge "Fixes OVN scenario"

6 years agoAdding NovaMigrationTarget service 89/55989/1
Ricardo Noriega [Wed, 18 Apr 2018 15:10:00 +0000 (17:10 +0200)]
Adding NovaMigrationTarget service

  Live migration fails due to this missing service
  in the compute role

JIRA: APEX-564

Change-Id: I13d69673204f6157dcbce31507aaa132f8c4ecce
Signed-off-by: Ricardo Noriega <rnoriega@redhat.com>
6 years agoFix nosdn fdio scenario 75/55775/5
Feng Pan [Sat, 14 Apr 2018 02:22:40 +0000 (22:22 -0400)]
Fix nosdn fdio scenario

- Update networking-vpp to use master branch
- Update THT vpp ml2 environment file with correct parameters
- Update external network creation command to use 'externa' as
  provider network name.
- Remove vpp network settings file as it's not used

apex-tripleo-heat-templates: Ia25db8456f1ad6beb96c7b9b5f318b166ef4576a
apex-puppet-tripleo: I231054a433eb7e598a6e24f6eaea02d476e776de

Change-Id: I4a1f68c75ae3b7d2a5b347d05abf0d025e8b116b
Signed-off-by: Feng Pan <fpan@redhat.com>
6 years agoFixes OVN scenario 19/55919/1
Tim Rozet [Tue, 17 Apr 2018 17:54:27 +0000 (13:54 -0400)]
Fixes OVN scenario

The OVN scenario would not deploy due to failures in trying to upgrades
to OVS 2.8 from OVS 2.7

JIRA: APEX-594.

Change-Id: Id84e488da8d2335f2240930c68119d0e2f6faf9c
Signed-off-by: Tim Rozet <trozet@redhat.com>
6 years agoFix barometer install failure 45/55745/1
Feng Pan [Sat, 14 Apr 2018 22:34:38 +0000 (18:34 -0400)]
Fix barometer install failure

We are seeing pip3 not found and collectd-write_sensu package conflicts
with latest pike overcloud images. This patch changes pip3 install to
python34-pip3 and updates collectd-write_sensu package version when
removing.

Change-Id: I7dfe1f2f39c3f21bffde8cfc6066f5bae66677d6
Signed-off-by: Feng Pan <fpan@redhat.com>
6 years agoEnable nosdn-fdio scenarios for Fraser release 53/54053/8
Feng Pan [Thu, 15 Mar 2018 23:44:39 +0000 (19:44 -0400)]
Enable nosdn-fdio scenarios for Fraser release

- Update neutron NSDriver patch
- Update os-net-config to fix config errors for vpp uio-driver
- Update puppet-tripleo to add configuration of physnets and type_driver
  in vpp agent
- Update THT to change VPP environment file for vpp-agent, and add common.yaml
  for vpp interface mapping
- Update VPP and networking-vpp version to 18.01
- Fix networking-vpp rpm build to enable proper uninstall.
- Update networking settings file to use ovs_bridge as default external interface type

JIRA: APEX-578
JIRA: APEX-568
JIRA: APEX-576
JIRA: APEX-577

apex-os-net-config: I915d5455acb8d496438b9c9e851639d3a43e6fa9
apex-puppet-tripleo: I472879b8f67e64b571638a0385943597a9120e6c
apex-tripleo-heat-templates: I5dfaf85d67fb038109edaf5c5d8a3e901b9148f4

Change-Id: I369bee232bfafef260d2ef19ac32614fdc487271
Signed-off-by: Feng Pan <fpan@redhat.com>