If we run script with -d switch and deployment status will timeout after
1500 seconds script will create diagnostic snapshot from env.
Change-Id: I84f480dc3f2ff33e1fc600c9a146cd8063b5a5a0
Some time ago there was decision to create directories without fuel-ccp
prefix. Now simple deployment job use builder that prepare directories
with this convention, so we should change this script to be compatible
with the other ones.
Change-Id: I1b1026b5a6ee851693ae1bc83112e0f44d49ce59
This patch changes names of networking interfaces used by CCP
mult-deploy script (and nothing else) to match defaults of
Ubuntu 16 VMs we build with packer.
It will provide more correct config for CI deployment jobs.
Change-Id: I06498b45090a7c9efc74853ee4b56f6bea232685
We include this file in ccp-configs-common.yaml so it should exist. We
generate content of this file in ccp-multi-deploy.sh script but sometimes we
may want to execute e.g. fetch command with this configs, so if we'll not
have this file it will be fail with 'no such file' msg like now our CI.
Change-Id: I64e029f9a8b8406694877ff2fa130462f72461ae
If APT_CACHE_SERVER variable is set then add option with it to config.
Example: APT_CACHE_SERVER="http://some_cache_server:3142"
Change-Id: Ic6d723d2715ab0be60a42f324b97d8ee02a1d97e
Disable multi-threaded build and increase timeout for registry
operations.
This patch only affects ccp-multi-deploy.sh Bash script (which is used
by CI) - it does _not_ change fuel-ccp defaults globally.
Change-Id: Ie02d55d8133b5a938ba2b3c06eb27f65c908a87d
This is required for kubectl call inside this script.
It's useful to properly run OpenStack tests on envs deployed in non-standard
K8s namespaces (which is the case for multiple deployment script).
Change-Id: Ia0076f881ca3d350e539bb4dd5d10f20484d8577
This may help with timeout errors we observer during build when
trying to push/pull large images.
--verbose argument to ccp should help in debugging failed builds.
Change-Id: I07a14d11a9cd5807dd52ce01931514ce273f163c
Network in OpenStack cannot be deleted if there are VMs using it.
VMs are being deleted in previous step of the script but it doesn't happen
synchronously unless using "--wait" argument.
Change-Id: Ia2e61e2602ac6816280d6aad2d9d1de15cd2a303
Since by default RabbitMQ uses a fixed nodePort we need to
change it in subsequent parallel deployments to avoid conflicts.
Change-Id: Ib1699136399e0a53d29a9d421ca83af12dceaf78
It's authored by Marek, I'm just reposting it here to make
publically available.
This is a sample shell script + config files to assist in deployment
of multiple OpenStack environments within a single K8s cluster.
Change-Id: Ic372d56b72b2430cabdb8afb84d2d7776a349465