Agbot container stuck in "restarting" status about devops HOT 10 CLOSED

@Al-tekreeti Is the arch amd64? Did the postgres container start in a healthy condition? I have this on my ubuntu 18 node, (NAME="Ubuntu" VERSION="18.04.5 LTS (Bionic Beaver)").

$docker ps |grep postgres
f1596ba35327   postgres:9                                    "docker-entrypoint.s…"   50 minutes ago   Up 50 minutes (healthy)   5432/tcp    

from devops.

@linggao Yes, all containers are healthy except Agbot. Sometimes vault container goes unhealthy, but if I restart the system using the flags -S and -s, it becomes healthy. The problem is only in Agbot container that keeps in a restarting state (does not reach unhealthy state even). It is ubuntu 20.04 not 18.04 as I mentioned earlier, sorry about that. Please see the following commands outputs:

$ uname -m
x86_64

$docker ps |grep postgres
3f462df198da postgres:latest "docker-entrypoint.s…" 4 days ago Up 9 minutes (healthy) 5432/tcp postgres

$ lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description: Ubuntu 20.04.3 LTS
Release: 20.04
Codename: focal

from devops.

@Al-tekreeti Can you show me how you start the all-in-one management hub? Your env variables for it?

from devops.

I forked the repo then I started it using the following command:
$ sudo ./deploy-mgmt-hub.sh

I don't pass any environment variables to it, but when the command finish execution it prints out the following tokens:

EXCHANGE_ROOT_PW=TEVUNLQzuQ3KNjYDfwveyEmjpnDZjU
EXCHANGE_HUB_ADMIN_PW=1W5VhfVxqja8e3hGYJyB8vepbWIsNO
EXCHANGE_SYSTEM_ADMIN_PW=qmbM5b6XAPI1cpHbLUBFMTPSVFv7lS
AGBOT_TOKEN=SMDCzgY8b8Irl7xdl5t6aAgYdcHp6F
EXCHANGE_USER_ADMIN_PW=Pe2UlU2RSinxeiDOp4fmEiXiAWNIB7
HZN_DEVICE_TOKEN=Z8Z3BbnDjlrUTYwZcEKGqSrRVJngkQ
VAULT_UNSEAL_KEY=BcbKa05euWxuWdkocZGvS2yijhvTumt236EFK1bloEE=
VAULT_ROOT_TOKEN=s.7zoHVzZNCIN0iCRzN9M1f0rg

from devops.

@Al-tekreeti I think it is the version of the postgres image (postgres:latest) that caused the problem. Please try to start the management hub with the following command for now.

export POSTGRES_IMAGE_TAG=9
sudo -sE ./deploy-mgmt-hub.sh
(where -E is used to preserve the env variable of the parent)

from devops.

@linggao Did not fix the issue and it still pulling postgres with "latest" tag.

from devops.

@Al-tekreeti That means the POSTGRES_IMAGE_TAG env variable is not taken. Please make sure you clean all the leftover containers before restart. (docker ps -a)
And please make sure to use the flag -sE for sudo command. Or you can use the following command:

sudo POSTGRES_IMAGE_TAG=9 ./deploy-mgmt-hub.sh

Anyway, find a way in your env to make the script take the POSTGRES_IMAGE_TAG env variable.
It seems that the latest release of postgres image broke the code.

from devops.

@linggao It did fix the issue. All containers now are healthy. Thanks a lot!

from devops.

@Al-tekreeti Good to hear. Btw, I tested POSTGRES_IMAGE_TAG=13, it works too. It is the version 14 which was released on 09/30/2021 that broke the code. We will stay with 13 for now. I will make a change in the deploy-mgmt-hub.sh to use 13 as default. Thanks for reporting this issue.

from devops.

You welcome! closing the issue.

from devops.