@bbjamm This is one of the reasons why I suggest the node.js install instead. Docker can be more complicated. In your case, the ./data folder was not owned by the user. So, in this case I suggest that you add your username to the docker group. The file ownership of files in the ./data folder will usually be the system ID. To change that, assuming user ubuntu:
sudo usermod -aG docker ubuntu
newgrp docker
groups
Another way to accomplish this is to find out your user id and your group id:
id -u
id -g
Then you could invoke docker-compose with the ID's:
docker-compose run -u 1001:1001 app id
Yet another way is to edit the docker-compose.yml file and include the UID/GID:
Awesome @vmsman ! You made this simple. Echo speaks is up and running locally. I just did the install on my 18.x ubuntu system directly, no docker. I don' t use it for much so this works for me.
I will be getting a UM350 after Christmas costs have been forgotten a bit by my wife
Note that I did have to reboot my server and restart the process manager manually due to the failed pm2 startup command. On my Synology box I apparently do not have any of the init system tools installed that pm2/lib/API/Startup.js is expecting to see, namely at least one of the following:
Which of these is the best to install to get the echo-speaks-server to run automatically on startup?
In spite of the restart issue, my local server appears to be running according to the status page at port :8091 and I have successfully reset the server info and re-logged into the Amazon authentication in the Echo Speaks app, which correctly picked up on my local server.
My server I moved from Herroku and all is running fine but it is at version 2.72. I see of these newer installs are at 2.8. What files do I need to update on the server to correct the version
For those who just stopped the app on heroku and removed their credit card number , be aware they still charged me 5 dollars since I had not removed the account.
Are you running all of this within a QNAP Container Station LXDs? I had mixed experiences with CS and various applications. PiHole ran fine, but Home Assistant was a nightmare (think it didnāt like the Network Bridging) until I gave up and installed it within Virtual Station. Problems immediately solved and it was within the same main network subnet.
I was thinking of just setting up a QNAP Virtual Station Ubuntu install then running your steps within that. Not sure if you have any thoughts or preferences regarding this and your preferred method. The NAS is probably beginning to hate me.
I am running the server in container station. You can search for the ātonesto7/echo-speaks-serverā image, configure the settings, and be up and running very quickly. See the Echo Speaks thread for further dialog on this.
I have run this and many other containers in the past on container station. I tend to recommend managing LXD containers from the QNAP command line. I do not like running docker instances directly inside of container station because of lack of management control. So, I tend to nest my docker containers inside of LXD. I have had as many as 36 containers running on my QNAP TS-1277.
EDIT: Google is my friend. Happened to search this and lo and behold, itās you:
So my understanding is simple and unfortunately Iām easily confused :
youāre running something like Ubuntu LXD within Container Station initially?
SSHing into that and installing docker directly.
Then within that running your containers independently (you wouldnāt be able so see all these within QNAPās CS main interface)?
Personally Iām liking Virtualization Station and would rather go that route with an Ubuntu install, but I donāt like how it uses more resources such as shared Kernels.
I have a TS-473a with QuTS after my 471 died recently (RIP almost 10 years.) Home office. I see youāre playing with the big kids on your 1277 lol. What the heck are you doing with all of that horsepower???
@leo_charles LXD containers use a lot less memory and CPU. I create a LXD container and then I install Docker inside of it and install a Docker container.
After messing with it I got it setup on Unraid with Portainer.
In Portainer
Add Container
Image Docker.io - tonesto7/echo-speaks-server:latest
Select the advanced options
add a Variable
Name - IPADDRESS
VALUE - ENTERYOURSERVERADDRESS
Deploy
Server should appear at your IPADDRESS:8091 in browser and in your unraid docker list
@vmsman has a better way to go about this but for simplicity's sake, I thought I'd explain a quick setup on my end us QNAP Container Station and local cookie management. I just don't have the time unfortunately to get into deeper weeds with alternate methods.
Steps:
Install Container Station on your QNAP if you haven't already.
\Create\ and enter eco-speaks into the search field (make sure it has the dash.)
Install Latest.
Keep default settings but set CPU Limit and Memory Limit to minimal.
Advanced Settings\Network\Network Mode\ set to Bridge.
Use Static IP (choose an IP that's available on your network.)
Create and wait.
Go to Container tab on left and then you might have to start the container that has echo-speaks.
Click on the echo speaks container. Should show you the URL (IP address you used) at the top with port 8091. Just remember this is the address.
head back to your ES install on HE and clear out the cookies field with heroku etc.
Add the local server option. You get a warning and a popup saying it's unsupported with a callback URL. Copy that callback URL as you will need to enter it on your Echo Speaks Service page and then save it there.
You'll want to reboot that container with the server, otherwise it will continue to say in HE that you need to log into amazon.
I'm sure I'm missing a few things, but if you got this far you can definitely figure it out. Oh and then run some device tests using ES. It's quite fast now.
UPDATE FIXED BELOW: It works, but the testing components to the actions are failing. Not sure why but hey Iāll take itā¦.
Have to add that I got the above working on the main page and with testing (speech, music, etc.) to all of my devices, but still canāt get any tests within actions to play any announcements. Perhaps these action tests arenāt working? Would love to know if anyone has any answers here.
Thereās one echo device which shows ānullā under the devices I activateānot sure if thatās normal but I canāt find any clue as to what it might have been (if anything at all.)
That's strange, I never had a CC on heroku, and after reading your post, I deleted the two apps I had and closed my acct. with no problem.
Wonder what's going on with them?