The code in docker/base/start.sh that waits for the log socket does
not work because it includes a bad "space" character after "-S".
This patch changes that character to a real "space" character. It
also sets the SKIP_LOG_SETUP envvar for the Heka container.
TrivialFix
Change-Id: I17a3005df5bbd57678350fc77030e44b5ad6acae
This is all kolla-mesos needs, so that we can run the extend_start
within the context of our start script. We need to do this to track
dependencies of the commands (one of which is the extend_start).
implements blueprint custom-extend-start
Change-Id: Ib844adc9c42dd8862281eb3f903d6c2b4695a34d
The extend_start.sh script for rsyslog is removed as it is no longer
needed. Docker no longer binds to /dev/log or /run/kolla/log
Closes-Bug: #1544545
Change-Id: Ic0a323a26ee4e9e15baf4598285844a8a4955f23
Part of ELK stack. Includes Dockerfiles for both Centos and Ubuntu.
Change-Id: I9f76adf084cd4f68e29326112b76ffd02b5adada
Partially-implements: blueprint central-logging-service
The install type is converted in kolla-build, so it will never
fail in dockerfile, move the check to kolla-build just above
the install type converting.
TrivialFix
Co-Authored-By: Jeffrey Zhang <jeffrey.zhang@99cloud.net>
Change-Id: I1500d3b47e909f94ea9f68c5245297733f63a70b
Our other keyservers for ubuntu use port 80 for firewall reasons.
Update the base to do the same.
TrivialFix
Change-Id: I4f6e59b3925e49a389e0415c943862cc13422f95
The base image is missing python-jinja2 for source builds which causes
the kolla_mesos_start script to fail on import.
Change-Id: I8550115dd42f4401a3351cd7c466fbeb1e02a665
Closes-Bug: #1532275
- Removed hardcoded yum repository configuration in favor of
commands dynamically generated based on repo-url and repo-file
arguments. We maintain a sane default set of repositories.
- Added generic rpm_setup_config parameter to add support for
installing .rpm or .repo files before building containers.
Co-Authored-By: Ryan Hallisey <rhallise@redhat.com>
Implements: blueprint custom-repos
Change-Id: I1b3a7647a9e7239de3cd162cb6f464f05632bde1
We want to record kolla version of running containers to be able to
detect whether or not we need to perform certain downtime-causing
actions during upgrade.
Change-Id: Ie113029da98303e6809d56edbf6d8de37be128d7
Implements: blueprint record-version
The ceph master repository is slow and the centos build often fails due
to timeouts fetching the ceph repo GPG key.
Switching to a more reliable mirror should improve things.
Change-Id: I7eef31fa9d83413a7c12134d285b3d20d95805e8
Closes-Bug: 1525505
