Some configuration options that with older openstack releases
were multistrings have now changed to csv values under a single
key, this change makes that simple to accomodate.
Change-Id: Id941a1e56e4a852d764084c958c13588b8e3ed85
Signed-off-by: Pete Birley <pete@port.direct>
In some instances the prometheus node-exporter provides inaccurate
feedback regarding disk usage. By adding the ability to add additional
mounts, more information about the node's filesystem can be made
available to monitoring services.
Change-Id: I5085a29683dc36099014efb1b11c7db774df501a
Co-Authored-By: Radhika Pai <rp592h@att.com>
Fixes an issue where dependencies are leaking from one resource to
another by ensuring that `envAll.Values.__dep` is reset each time.
Change-Id: I34a4b2fa70d608b2c69bdf18275e439f96976229
Closes-Bug: #1845538
This change adds default network policy definitions for ingress/egress
in the prometheus-openstack-exporter chart to allow all traffic.
This also adds a netpol value override to enable network policy for
various network policy checks.
Change-Id: If1314420d5038174e8641c9809de65b23853d3f3
Added the reblabeling config lines to the kubernetes_sd_config key, to
replace the node_name with hostname for Node-exporter. This must now
display the hostname also as one of the labels of the Node-exporter
metrics.
Change-Id: Ic96a890552a1cd2f5e595c37330de048f31a0e75
This PS updates the mariadb ingress error page server to run as the
nobody user.
Change-Id: I13756ba79e8c7b857e0807447192e06b11762abf
Signed-off-by: Pete Birley <pete@port.direct>
This is to fix the logic to find empty directory which has lost+found
folder created by linux filesystem.
This is giving problem when linux partition mounted and using it as file
backed osd.
Change-Id: I11e68a7b9e5df6657d8bf199b8030bc3fc51595e
This change adds network policy overrides for multiple infra
services for the openstack-helm network policy gate.
Change-Id: If051ec1749cb9ed1e289f0cf82a8876371e36531
The elasticsearch dashboard was not showing correct value for the
unassigned shards panel. Updated the query.
Change-Id: I60056daede2957a45958769dd6b723f54cc54ed7
This mop covers how to move single osd chart which is already installed
to multiple ceph osd charts. This will give more flexibility to handle
upgrade activites on ceph cluster .
Change-Id: Ib5f8c5c38df296e4783f2af43c6dc4e769fe4140
This adds BlueStore support for the ceph-osd chart so that OSDs
may be deployed using BlueStore with optional --block.db and
--block.wal parameters.
Co-Authored-By: Chinasubbareddy Mallavarapu <chinasubba.reddy@att.com>
Change-Id: Ifbae8331b595c15c168ccd6e93b00ff054a607bc
This patch set adds the ability to add a label to the following namespaces:
- kube-public
- kube-system
- default
This was previously done in kubeadm in patch [0] but appears to be
removed as the script moved around. Also, this PS adds this to the
minikube deployment.
[0] https://review.opendev.org/#/c/540131/
Change-Id: I0f06e8ae0cd7742313b447dc2d563c7d92318fb0
Signed-off-by: Tin Lam <tin@irrational.io>
Setup Cgroup to use to break out of Kubernetes defined groups for ovs-dpdk pods.
All the cores on the server are added to the cpuset, pmd_cpu_mask and lcore_mask
will choose the right ones for ovs-dpdk from all the cores.
Co-Authored-By: Phil Sphicas <ps3910@att.com>
Change-Id: Ia840647e3fc09480b826b3075b2585daefa638b3
This change adds egress rules to the following charts:
- ingress
- memcache
- libvirt
- rabbitmq
These rules will be tightend down in future changes
Change-Id: I6f297d50ca4c06234c7c79986a12cccf3beb5efb
This removes a duplicate rules: key that was erroneously added
to the elastic-filebeat/metricbeat and fluentbit daemonsets
Change-Id: Ic2a6347ae69ccbd2b9075d9ee0180ad41932a9bf
Signed-off-by: Steve Wilkerson <sw5822@att.com>
This adds support for configuring fluentd's update strategy when
deployed as a daemonset, as this was previously missed when
the changes to support both daemonsets and deployments were made
Change-Id: I5ac4fbfc0e64caaf207de42cd71c893f8d0f6ff1
Signed-off-by: Steve Wilkerson <sw5822@att.com>
This updates the Fluentd deployment template to use the helm
toolkit util for generating environment variables through the
chart's values.yaml. This adds flexibility in defining fluentd
outputs, as arbitrary environment variables can be injected and
consumed in fluentd's filters and outputs
Change-Id: I72a2c476378cc555bde1387781b4a06f13b51bc6
Signed-off-by: Steve Wilkerson <sw5822@att.com>
This updates the fluentd chart to add support for leveraging a
Kafka output. This required adding a kafka endpoint entry to the
chart's values.yaml, as well as the required template updates to
the fluentd deployment template and the addition of a secret for
kafka credentials
Depends-On: https://review.opendev.org/#/c/679297/
Change-Id: I80a487a0538f0b3704fb598da38c07feedaccb0e
Signed-off-by: Steve Wilkerson <sw5822@att.com>
This updates the grafana and nagios helm test pod templates to
use the internal endpoints for their selenium tests instead of the
public endpoints when defined
Change-Id: I1138cb29a808894d3339bc1b07c3a60804b9546f
Signed-off-by: Steve Wilkerson <sw5822@att.com>
This PS adds the ability to attach a release uuid to pods objects.
Implements: Ability to attach release_uuid to ephemeral pods
Change-Id: I0ebade75e18eced99fe16ba434558264b1793e88
This is to fix name conflict for configmap name "ceph-osd-default" when
we try to create multiple osd releases as every relase try to create configmap
with same name.
we could add relase name here but that will be a problem for sites deployed with
current logic as upgrade will delete old daemonsets and creates new ds ,so all osd
pods gets recreated at a time, by getting this from values can give us
flexibility to install multiple osd releases with out effecting currently deployed
sites.
Here is the error if we try multiple osd releases with current logic:
2019-08-27 13:54:16.690 41 ERROR armada.handlers.tiller [-] [chart=ceph-osd-sde]:
Error while installing release ceph-osd-sde: grpc._channel._Rendezvous: \
<_Rendezvous of RPC that terminated with:
status = StatusCode.UNKNOWN
details = "release ceph-osd-sde
failed: configmaps "ceph-osd-default" already exists"
Change-Id: Ibe84582b9ba04c6cbf611e943ecd0a7149c5ab2f
