Token expires after 24h and is not renewed

[vrachieru]

The Skype token is valid for 24h after its creation and it is not properly renewed via SkypeConnection#refreshSkypeToken.
More info can be found in this and this issue.

As a temporary solution, the container can be restarted as follows:

*/5 * * * * /srv/ursar-bot/restart.sh

skype_username='username'
skype_password='password'
container_name='ursar'

is_running=$(docker exec ursar echo true 2>/dev/null || echo false)

echo "Is running: $is_running"

if [ "$is_running" = "true" ]
then
  start=$(docker inspect -f {{.State.StartedAt}} "$container_name")
  start_ts=$(date -d "$start" '+%s')
  now_ts=$(date '+%s')
  running_hours=$(( ( now_ts - start_ts ) / (60*60) ))
  echo "Running hours: $running_hours"
fi

if [ "$is_running" = "false" ]
then
  echo "Starting $container_name container"
  docker run --rm -t -e URSAR_SKYPE_USERNAME="$skype_username" -e URSAR_SKYPE_PASSWORD="$skype_password" --name "$container_name" academiadeursarie/ursar-bot
fi

if [ "$is_running" = "true" ] && [ "$running_hours" -ge "24" ]
then
  echo "Restarting $container_name container"
  docker stop "$container_name" && docker run --rm -t -e URSAR_SKYPE_USERNAME="$skype_username" -e URSAR_SKYPE_PASSWORD="$skype_password" --name "$container_name" academiadeursarie/ursar-bot
fi

[vrachieru]

After further investigation it might be sufficient to provide a file in the SkypeEventLoop constructor where the token would be saved and then used to create a new token.
The current issue being that the token is not persisted anywhere.

Example:

# https://github.com/academia-de-ursarie/ursar-bot/blob/master/ursar/main.py#L14
super(UrsarBot, self).__init__(settings['SKYPE_USERNAME'], settings['SKYPE_PASSWORD'], 'token.txt)

[fahadrokomari]

Is this issue fixed or any way around with this?

[vrachieru]

Read the explanation above...
It's not fixed, as you can see, the issue is still open.

[fahadrokomari]

@vrachieru Ok... Is there any way around this ?

[vrachieru]

@fahadrokomari It's explained above how you can restart the container with a cron to work around this issue..
The script is literally the first thing in this issue.
WTF?!

[fahadrokomari]

WTF= Wow that's fun.
Yes, I have got this. I do restart the script. But it seems not very legit. Thank you for your help though. I was doing this for a while but thought there is some other way around.

Restarting a container is like a "Microsofts" solution. By the way, have a nice day.

[vrachieru]

@fahadrokomari Well that's what a workaround is, it's not pretty, I agree 😄 .
If you do find a better way PRs are welcome.
As you can see we haven't touched this project in a while so fixing this isn't a priority.
You have a nice day too.