Acumos Boreas install fails starting docker-proxy on Ubuntu 18.04 VM

Question

I am now trying to install Acumos Boreas on an Ubuntu Quemu/KVM VM with 12CPU's (cores) 32G RAM and 300G disk space following this procedure:

0) git clone https://gerrit.acumos.org/r/system-integration

1) bash setup_docker.sh

Add user to docker group: sudo usermod -aG docker $USER and logout/in again

2) bash system-integration/AIO/acumos_k8s_prep.sh $USER acumos 2>&1 | tee aio_prep.log

3) cd system-integration/AIO

4) bash oneclick_deploy.sh

It all looks good (pods are however taking very long to get out of pending state) until the install procedure tries to install the docker-proxy. The docker-proxy pod does not get out of pending state in time:

.
.
.

check_running:357 (Mon Jul 22 14:15:30 UTC 2019) docker-proxy status is Pending
+ [[ Pending != \R\u\n\n\i\n\g ]]
+ [[ 31 -le 30 ]]
+ [[ 31 -gt 30 ]]
+ [[ k8s == \d\o\c\k\e\r ]]
+ inspect_pods_for_app docker-proxy acumos
+ trap fail ERR
+ local app=docker-proxy
+ local namespace=acumos
++ kubectl get pods -n acumos -l app=docker-proxy -o json

.
.
.


+ name='"docker-proxy"'
+ kubectl logs -n acumos -l app=docker-proxy -c '"docker-proxy"'
Error from server (BadRequest): container "docker-proxy" is not valid for pod docker-proxy-595678b865-2qpg2

Any ideas? To me it seems strange that the pods takes so much time to get started (pending) given that the VM has 12 cores and I wonder if the kubernetes installation is managing to utilise the full range of available CPU cores?

Answer 1

thanks for trying this out. I'll help you resolve the issue. Note the current release is still a WIP and while these scripts are regularly tested in our test envs, variations outside that can result in issues. Resolving these issues helps us improve the broader usability/reliability of the platform deployment, so again thanks for making the effort.

Usually when a pod takes too long to become active there is some underlying issue. You can get more information from: kubectl describe pods -n acumos -l app=docker-proxy kubectl logs -n acumos $(kubectl get pods -n acumos -l app=docker-proxy | awk '/docker-proxy/{print $1}')

If you see something that looks like an error or unmet constraint in there, let me know. Sometimes however there are issues with upstream dependencies not being available at the moment, eg for software that needs to be installed, so just trying the deployment again will often resolve the issue unless there is some specific error happening.

Answer 2

好的，所以我在utils.sh中将30更改为300（在第99、311、392和397行中），在setup_nexus.sh中将60更改为600（在第119行中），从而设法安装了Acumos：-）...似乎可以正常运行，但稍后必须更详细地进行检查。

Answer 3

Looking at the output from the suggested kubectl commands (and info in the Kubernetes Dashboard) I have observed the following:

1) I see a lot of this : Pod The node was low on resource: [DiskPressure] , which is strange since the node has plenty free diskspace and inodes.

2) The docker-proxy and docker-dind service fails to start the associated pod multiple times, the docker-proxy finally "succeeds" to run but errors in logfile (see below).

3) The docker-dind-xxxx pod fails completely to start up : State: Waiting, Reason: CrashLoopBackOff, Last State: Terminated, Reason: Error

In the logfile for the docker proxy I see this:

Reading state information...
E: Unable to locate package python
E: Unable to locate package python-pip
Waiting for temporary DNS resolution issue to be fixed
Reading package lists...
Building dependency tree...
Reading state information...
E: Unable to locate package python
E: Unable to locate package python-pip
Waiting for temporary DNS resolution issue to be fixed
Reading package lists...
Building dependency tree...
Reading state information...
E: Unable to locate package python
E: Unable to locate package python-pip
Waiting for temporary DNS resolution issue to be fixed
Reading package lists...
Building dependency tree...
Reading state information...
E: Unable to locate package python
E: Unable to locate package python-pip

Any ideas on how to solve this?