This commit creates cgts-vg volume group automatically on worker
nodes by kickstart. This cgts-vg volume group reserves space for
log-lv, scratch-lv, docker-lv and ceph-mon-lv.
This commit reserves space in cgts-vg volume group for 30G
docker-lv and 20G ceph-mon-lv for AIO configuration.
Story: 2004520
Task: 28663
Change-Id: Ic77d00c354da1070e2c4c2da4545d70ab4a93d91
Signed-off-by: Wei Zhou <wei.zhou@windriver.com>
Starting collectd too early in the manifest apply is seen
to occasionally fail due to a dependency configuration on
hostname resolution in FQDNLookup not being complete.
Since influxdb is used by collectd and is a controller
only service this update moves it to the manifest apply
post stage as well and is filtered out from non
controller load types.
This issue is fixed by the following multi-git changes.
stx-metal: This update.
Filter influxdb out of storage and compute only loads.
No real inter git merge dependency
stx-integ:
Add startup Before=pmond dependency
stx-config:
Move collectd config and startup to manifest apply post stage
Move influxdb config and startup to manifest apply post stage
Test Plan:
PASS: Build iso
PASS: verify install storage system and collectd startup
PASS: Verify Storage system DOR
PASS: Verify influxdb and extensions excluded in non-controller loads
PASS: Verify collectd starts properly on all nodes (CC,DOR,UNLOCK)
PASS: Verify influxdb starts properly on controller nodes (CC,DOR,UNLOCK)
PASS: Verify collectd pmond process monitoring and recovery
PASS: Verify influxdb pmond process monitoring and recovery
PEND: Verify collectd statistics storage and fetch to/from influxdb
PEND: Install AIO DX and verify collectd and influxdb startup
Change-Id: I8c71f36978620e0650062cc848bfb9d85f6810b2
Closes-Bug: 1797909
Signed-off-by: Eric MacDonald <eric.macdonald@windriver.com>
This increases the default docker distribution partition size from
1G to 16G. This also increases the minimum disk requirements from
130G to 145G for small disk, 170G to 185G for large disk.
Story: 2004520
Task: 28526
Change-Id: I898cfac45757ff1f9e6ce7c4928bbd9a42dca77d
Signed-off-by: Angie Wang <angie.wang@windriver.com>
This update replaces compute references to worker in mtce,
kickstarts, installer and bsp files.
Tests Performed:
Non-containerized deployment
AIO-SX: Sanity and Nightly automated test suite
AIO-DX: Sanity and Nightly automated test suite
2+2 System: Sanity and Nightly automated test suite
2+2 System: Horizon Patch Orchestration
Kubernetes deployment:
AIO-SX: Create, delete, reboot and rebuild instances
2+2+2 System: worker nodes are unlock enable and no alarms
Story: 2004022
Task: 27013
Depends-On: https://review.openstack.org/#/c/624452/
Change-Id: I225f7d7143d841f80459603b27b95ac3f846c46f
Signed-off-by: Tao Liu <tao.liu@windriver.com>
Use openstack-aodh-config package to package service and config
files for openstack-aodh package.
The openstack-aodh-config need to be set to the same node filter
as we did for openstack-aodh.
Deployment test pass and related service/script file check pass.
Story: 2003768
Task: 28044
Change-Id: I454762abc1dbd6c5db639be2b1f046e23c131d91
Signed-off-by: zhipengl <zhipengs.liu@intel.com>
This commits add the creation of a symlink for EFI/grub.cfg. It also
allows the user to overwrite an existing directory, the user will be
prompted to confirm this action.
Closes-Bug: 1794863
Change-Id: I566a3b39c601921cd73cca3f291f43e5dc0ef626
Signed-off-by: Kristine Bujold <kristine.bujold@windriver.com>
Barbican is going to be running on controller node only.
So, I'm removing all Barbican rpms from compute and storage nodes.
Change-Id: Ib00d697a2c9816cc7c3f181bb0f4d298bba973bd
Story: 2003108
Task: 27700
Signed-off-by: Alex Kozyrev <alex.kozyrev@windriver.com>
Use net-snmp-config package to package script and service file for
net-snmp-config package.
Like net-snmp, net-snmp-config will also not be installed on computer
and storage node.
Deployment test and ping test between VMs pass
Config,serivce and script files check pass.
Story: 2003768
Task: 27586
Change-Id: I2f9d1bfdbe0b27fbd58137df8a3fd36d3053defa
Signed-off-by: zhipengl <zhipengs.liu@intel.com>
Base disk increases by 6G from 124G to 130G for
small disk and from 164 to 170 for large.
Story: 2002876
Task: 27948
Change-Id: I3d987c0d70bf18a91cb4c977ac16fcdabe2cb9fc
Signed-off-by: Al Bailey <Al.Bailey@windriver.com>
Package openldap-config is added to config customized config file
of openldap. These config files were packaged to openldap-servers,
so add the filter the same as openldap-servers.
Story: 2003768
Task: 26462
Change-Id: Id40967bcbed40998649602c209e8608532584058
Signed-off-by: slin14 <shuicheng.lin@intel.com>
Package lighttpd-config is added to config customized config
file of lighttpd. Add the filter the same as lighttpd.
Story: 2003768
Task: 27616
Change-Id: Ib221c12cdd5a6ee88ceb3666ae210a4180c621b4
Signed-off-by: slin14 <shuicheng.lin@intel.com>
openvswitch-config is created for openvswitch config file
configuration. Add it to the filter the same as openvswitch.
openstack-neutron-openvswitch is filtered from controller node.
It is a Neutron plugin that implements virtual networks using
Open vSwitch. Should be for compute node only.
Otherwise, due to it depends on openvswitch, it will causes
openvswitch is installed in controller node also.
Test:
Pass build and multi node deploy test.
Story: 2003768
Task: 27601
Change-Id: Ieaaa1e5db31bca5262cbbcd61ebfc0390be10528
Signed-off-by: slin14 <shuicheng.lin@intel.com>
This change filters out the new package from non-controller nodes.
Depends-On: https://review.openstack.org/615923
Change-Id: I6b9c078c19bf37ef0312947fc2a4249347a604f7
Signed-off-by: Jason McKenna <jason.mckenna@windriver.com>
Story: 2004240
Task: 27763
Use haproxy-config package to package config files for haproxy package
Haproxy & haproxy-config will not be installed on storage node.
Deployment test pass and ping pass between 2 VMs.
Script file and service file have been checked in system folder.
Story: 2003768
Task: 27584
Change-Id: I026697079f968b54b8c518abeab847e95dc79e49
Signed-off-by: zhipengl <zhipengs.liu@intel.com>
Need to change related filter files as we only install
memcached package in controller node.
Basical multinode deployment test pass.
Story: 2004108
Task: 27517
Change-Id: Id21d3db9d172398f6c23f86e5dd5f9e5a249c6b8
Signed-off-by: zhipengl <zhipengs.liu@intel.com>
Rename files and folders in mtce-compute, mtce-control, and
mtce-storage. As well update packages' names in bsp-files/
filter_out_* scripts accordingly.
Story: 2004079
Task: 27485
Change-Id: Ic1e9bd4bb8d72f30ddcc2a2bfc602a1a34e583da
Signed-off-by: Yong Hu <yong.hu@intel.com>
Make sure the minimally sized physical volume for the cgts-vg volume
group is sized correctly when the --kubernetes option is used with
config_controller. Include the kubernetes specific logical volumes that
will be present in a kubernetes configuration.
Change-Id: Ifc6d2c0a5dbb880f8e1ff73b01bc05cb8ab22855
Partial-Bug: #1794567
Signed-off-by: Robert Church <robert.church@windriver.com>
The UEFI grub config file used for PXE Boot Server containes a mixed
of menuentry and submenu. A submenu opens a new context and global
variables cannot be used within them
https://bugs.launchpad.net/ubuntu/+source/grub2/+bug/1175127. This
was causing pxe_root to be empty for submenus selections. Moving
'1) UEFI Boot from hard drive' into a submenu fixes this problem and
also removes the need to save the root variable. This was previously
required when a user would select boot from hard drive and the hard
drive did not exist, root would no longer be set to the pxeboot
server and selecting the other menu entries would fail.
Remove the '1)' prefix in the 'Boot from hard drive' menu option in
both Legacy and UEFI boot as this is not used by other menues.
Closes-Bug: 1794863
Change-Id: I5bc62039bfb68477e9cd0166ce17877693037640
Signed-off-by: Kristine Bujold <kristine.bujold@windriver.com>
Minimum backup partition size is calculated as the sum between the
sizes of the glance and database partitions, as well as a 20GB
overhead.
This fix increase the default backup size partition to 40GB to
be in line with the above calculations, considering the database
and the glance partitions are 10GB each by default.
This also increases the minimum disk requirements from 120GB
to 130GB.
Change-Id: I5cfc329870a84a6245d868b4c4990829e702e886
Closes-bug: 1793543
Signed-off-by: Stefan Dinescu <stefan.dinescu@windriver.com>
Allocating of space on the root disk was done in kickstart, combined
with config_controller. Various checks for disk and volume group
space made it difficult to add new filesystem partitions.
To simplify this, all checks are now merged under disk size checks.
Kickstart files now create the partitions of the same size for
rootfs and log partitions, no matter the disk size, with the
only variable size being the cgts-vg size.
Thus, any future changes will only need to consider only this
size for ensuring enough space is present.
The limit between small and big disks is 240GB, with a minimum
disk size of 120GB.
Depends-On: https://review.openstack.org/#/c/600743/
Change-Id: I37ecc8eb5468811d1ca3a71f8e2a0629525e8fad
Closes-bug: 1791170
Signed-off-by: Stefan Dinescu <stefan.dinescu@windriver.com>
Filter out platform-util-controller from compute and storage nodes.
Story: 2002826
Task: 26228
Depends-On: https://review.openstack.org/600868
Change-Id: I3d8fc737c9a59caef7658d4139b302cee3841592
Signed-off-by: Jack Ding <jack.ding@windriver.com>
use flake8 as pep8 tools
enable check and gate for pep8(voting)
Fix below flake8 issues:
E127 continuation line over-indented for visual indent
E211 whitespace before '('
E222 multiple spaces after operator
E302 expected 2 blank lines, found 1
E501 line too long (101 > 79 characters)
E502 the backslash is redundant between brackets
F401 'platform' imported but unused
W391 blank line at end of file
Change-Id: Idfb953e52c8ee35c2adefdf0e4143a381c7f49e2
Story: 2003426
Task: 24596
Signed-off-by: Sun Austin <austin.sun@intel.com>
Fix below linters issues
E001 Trailing Whitespace
E003 Indent not multiple of 4
E006 Line too long
E011 Then keyword is not on same line as if or elif keyword
E020 Function declaration not in format ^function name {$
E040 Syntax error: syntax error near unexpected token `;'
ignore cases are added in tox setup
E006 Line too long
E010: do not on the same line as for
Story: 2003368
Task: 24427
Change-Id: I6acf64271a4e608be8bc8fa965cac4fa31e0c05b
Signed-off-by: Sun Austin <austin.sun@intel.com>
This is an enhancement request to add the screen package to
controller nodes
This specific modification prevents the screen package from being installed
on other nodes (compute or storage)
The screen package is added in another commit
(see https://review.openstack.org/#/c/595249/)
Story: 2003061
Task: 23100
Depends-on: https://review.openstack.org/#/c/595249/
Change-Id: I355d517ba0d0392d40fe78991798ddf6e5d16fde
Signed-off-by: Paul-Emile Element <Paul-Emile.Element@windriver.com>
The low-capacity Swift solution this Story is implementing is on
controllers only.
Story: 2003518
Task: 24811
Depends-On: https://review.openstack.org/595330
Change-Id: I7bb98195bbda2a97f004329f024701475f139d53
Signed-off-by: Jack Ding <jack.ding@windriver.com>
Filter out the fm client and fm rest api packages from
compute and storage nodes
Story: 2002828
Task: 22747
Depends-On: https://review.openstack.org/#/c/591452/
Change-Id: If0663dfb2cc1b557a1b9439c64d3ccb36bd66503
Signed-off-by: Tao Liu <tao.liu@windriver.com>
Updating kickstart to provision 5G for new gnocchi filesystem in
cgcs disk partition.
Story: 2002825
Task: 24240
Change-Id: Ie6182a636e6b9c580af2cce671dcbb267acb305f
Signed-off-by: Angie Wang <angie.wang@windriver.com>
To improve kubernetes support, update kernel to CentOS 7.5 version
and enable user namespaces in kernel bootargs.
Depends-On: https://review.openstack.org/580689
Change-Id: I4d8620ea17a19a764c6627cd79eb548c79c56bfd
Signed-off-by: Jason McKenna <jason.mckenna@windriver.com>
Story: 2002761
Task: 22841
Implements customer configuration of kernel options to control
spectre/meltdown related kernel options. Default (with "nopti
nospectre_v2" options) can be changed to "" using
system modify -S spectre_meltdown_all
Change-Id: I183a22fa681e6524415558c0009aa8786418cc07
Signed-off-by: Jack Ding <jack.ding@windriver.com>
When the pxeboot setup feature is used from a patched ISO, the patch
repo is downloaded from the pxeboot server during installation.
The kickstart was missing a change in working directory when mirroring
the patch repo, leading to issues when installing the second controller.
This update corrects this issue.
Change-Id: I5926ee4f196adf3938b8934f57c15eadde83a5fb
Signed-off-by: Don Penney <don.penney@windriver.com>
To avoid potential issues due to large time jumps when NTP first syncs
the system time at runtime, this update adds ntpd to the installer
rootfs and adds a pre-script to the kickstarts to sync the time from
the active controller before starting to install the software. This
also ensures that any filesystem timestamps will be accurate right
from the node installation.
Change-Id: I166c52430cec6ba64e5a33ebde64ee65639d623c
Signed-off-by: Don Penney <don.penney@windriver.com>
During patchback testing, it was found Anaconda may also install
mlnx-ofa_kernel-rt from the patching repo, if it's in the patch.
This update adds it to the exclusion lists.
Change-Id: I0809898c2dc748543033bc6b11d5e22bd8e462ad
Signed-off-by: Don Penney <don.penney@windriver.com>
The prepatching feature changed how kickstarts are delivered. Where
before, they were generated as part of build-iso and copied during
installation, they are now delivered in packages. However, when the
controller kickstart mirrors the feed directory from the active
controller after installation, it may overwrite these installed files.
This update changes the wget command options to protect against this.
In addition, testing with an RT kernel patch showed Anaconda was also
installing the mlnx-ofa_kernel-rt-modules package from the patching
repo on a standard note, as it attempts to resolve a packaging
requirement. This update also adds explicit exclusions to the package
lists in the standard and lowlatency kickstarts to avoid installing
rt modules on the standard nodes, and vice versa.
Change-Id: I56b22fb0846db05a96004184c1060c05566d5363
Signed-off-by: Don Penney <don.penney@windriver.com>
Filter out memcached service on compute and storage nodes.
Change-Id: Ie558ac297d4943038ecb194b38cd4825afc61f99
Signed-off-by: Jack Ding <jack.ding@windriver.com>