Errors trying to install Acumos

Question

Background： VMware15.0 ubuntu16.04-64bit 32G RAM + 16Core CPU /etc/hosts: 192.168.79.130 localhost

Doing this (and typing the sudo password when prompted):

git clone https://gerrit.acumos.org/r/system-integration
apt-get -y update
apt-get -y install docker-ce=18.06.3~ce~3-0~ubuntu
if [[ "$(id -nG "$USER" | grep docker)" == "" ]]; then sudo usermod -aG docker $USER; fi
# Logged out and in again and verified that my user is in the docker group
cd system-integration/tools/
bash setup_k8s_stack.sh setup
cd
bash system-integration/AIO/setup_prereqs.sh k8s localhost $USER generic 2>&1 | tee aio_prep.log
# When "Prerequisites setup is complete" messages is displayed I continue with
cd system-integration/AIO
bash oneclick_deploy.sh 2>&1 | tee aio_deploy.log

The deployment fails with the following error message:

+ c='-l component=192.168.79.130'
++ kubectl get deployment -n acumos -l app=cds -l component=192.168.79.130 -o json
++ jq -r '.items[0].metadata.name'
+ dep=null
++ cat /tmp/a72a447b-df96-4fec-98c9-bb99e447d00d
+ kubectl patch deployment -n acumos null --patch 'spec:
  template:
    spec:
      hostAliases:
      - ip: "192.168.79.130"
        hostnames:
        - "ubuntu"'
Error from server (NotFound): deployments.extensions "null" not found

Open file：“system-integration/AIO/utils.sh”

  if [[ "$component" != "" ]]; then c="-l component=$component"; fi
  dep=$(kubectl get deployment -n $namespace -l app=$app $c -o json | jq -r ".items[0].metadata.name")
  kubectl patch deployment -n $namespace $dep --patch "$(cat $tmp)"

Modified to：

  #if [[ "$component" != "" ]]; then c="-l component=$component"; fi
  dep=$(kubectl get deployment -n $namespace -l app=$app $c -o json | jq -r ".items[0].metadata.name")
  kubectl patch deployment -n $namespace filebeat --patch "$(cat $tmp)"

This error was resolved, but the following error occurred:

oneclick_deploy.sh setup_federation:233 (Tue Sep 17 02:32:31 PDT 2019) CDS API is not yet ready; waiting 10 seconds
+ t=300
+ sleep 10
++ curl -k -u ccds_client:187bbf19-40b9-45c8-9945-4903292d963d https://localhost/ccds/peer
++ grep -c numberOfElements
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100   203  100   203    0     0    957      0 --:--:-- --:--:-- --:--:--   962
+ [[ 0 -eq 0 ]]
+ [[ 300 -eq 300 ]]
+ fail 'CDS API is not ready after 300 seconds'
+ set +x

oneclick_deploy.sh fail:42 (Tue Sep 17 02:32:41 PDT 2019) CDS API is not ready after 300 seconds

pod cds has occurred an error: CrashLoopBackOff logs:

Events:
  Type     Reason     Age                  From               Message
  ----     ------     ----                 ----               -------
  Normal   Scheduled  13m                  default-scheduler  Successfully assigned acumos/cds-7474fccbc7-nm4jk to ubuntu
  Normal   Pulled     9m28s (x5 over 13m)  kubelet, ubuntu    Container image "nexus3.acumos.org:10002/common-dataservice:2.2.5" already present on machine
  Normal   Created    9m28s (x5 over 13m)  kubelet, ubuntu    Created container
  Normal   Started    9m28s (x5 over 13m)  kubelet, ubuntu    Started container
  Warning  BackOff    3m9s (x28 over 12m)  kubelet, ubuntu    Back-off restarting failed container

This problem is similar to the previous one： Errors trying to install Acumos Boreas release Thanks！

logs:('Patch deployment for cds')

...
oneclick_deploy.sh start_deployment:518 (Tue Sep 17 18:25:16 PDT 2019) Creating deployment cds
+ kubectl create -f deploy/cds-deployment.yaml
deployment.apps/cds created
+ get_host_ip_from_etc_hosts localhost
+ trap fail ERR
++ grep -v '^127\.'
++ awk '{print $1}'
++ grep -E '\slocalhost( |$)' /etc/hosts
+ HOST_IP='192.168.79.130
192.168.79.130'
+ [[ 192.168.79.130
192.168.79.130 != '' ]]
+ patch_deployment_with_host_alias acumos cds ubuntu 192.168.79.130 192.168.79.130
+ trap fail ERR
+ namespace=acumos
+ app=cds
+ name=ubuntu
+ ip=192.168.79.130
+ component=192.168.79.130
+ log 'Patch deployment for cds (192.168.79.130), to restart it with the changes'
+ setx=x
+ set +x

oneclick_deploy.sh patch_deployment_with_host_alias:448 (Tue Sep 17 18:25:17 PDT 2019) Patch deployment for cds (192.168.79.130), to restart it with the changes
++ uuidgen
+ tmp=/tmp/ef6cab06-ece6-436b-812c-1a00728bec01
+ cat
+ [[ 192.168.79.130 != '' ]]
+ c='-l component=192.168.79.130'
++ kubectl get deployment -n acumos -l app=cds -l component=192.168.79.130 -o json
...

Answer 1

Two suggestions:

(1) Add yourself to password-less sudo permissions

sudo visudo
(add to the end of the file and save)
<your username>   ALL=(ALL:ALL) NOPASSWD:ALL

2) Re 'The deployment fails with the following error message:', '-l component=192.168.79.130' indicates a bug that caused patch_deployment_with_host_alias (in utils.sh) to think there was a component specified (parameter 5). The specific call to this was from start_acumos_core_app for the cds app, in line

patch_deployment_with_host_alias $ACUMOS_NAMESPACE $app $ACUMOS_MARIADB_HOST $HOST_IP  
or
patch_deployment_with_host_alias $ACUMOS_NAMESPACE $app $ACUMOS_HOST $HOST_IP

Since there are at most four parameters in that (as of master), can you provide a little more of the log (back to the line that was output: 'Patch deployment for cds').

Getting beyond that did not require updates to utils.sh and I think the changes you made had side-effects probably. So I would reverse those changes: - you replaced $dep with filebeat This may have allowed you to progress but it broke the purpose of that function (to add a host alias for components - not just filebeat - that refer to names not resolvable in DNS).