SMS desktop app

[timconsidine]

@privsec it does take a moment or two before starting up
but if this persists, seems like it is not finding the app

at the risk of basics for troubleshooting:

• inside the container and inside the installation directory, you have done node app.js & to start it (I have forgotten a couple of times). I think you have because you got database connected but maybe the program stopped for some other change

• outside the container, you have run systemctl reload nginx without errors ? and/or run nginx -t without errors ?

• these errors suggest continuing nginx conf issue

2021/11/03 13:22:47 [warn] 48507#48507: conflicting server name "app.domain.com" on 0.0.0.0:80, ignored
2021/11/03 13:22:47 [warn] 48507#48507: conflicting server name "app.domain.com" on [::]:80, ignored
2021/11/03 13:23:14 [warn] 48554#48554: conflicting server name "app.domain.com" on 0.0.0.0:80, ignored
2021/11/03 13:23:14 [warn] 48554#48554: conflicting server name "app.domain.com" on [::]:80, ignored

what is latest / current nginx conf file ?

[timconsidine]

@privsec said in SMS desktop app:

@privsec
I have confirmed that the all three nginx files use the LXC IP address, not the VPS IP address.

maybe I am having a brain outage, but why are there 3 nginx files ?

[privsec]

@timconsidine

Well, I’m now unable to load the server. It seems that after the reboot I requested it is just frozen at the boot stage.

So I’m gonna have to start over.

However, I kept running into errors with certbot originally, until I made a .wellknown folder

[privsec]

@timconsidine are there tried and true steps you do to setup certbot with ngoni every time?

[timconsidine]

@privsec all those difficulties doesn't seem right to me

What OS is on the VPS and what OS did you choose for the container ? (FYI on my installation it is Ubuntu 20.04 in both)

Given that we are dealing with a container here (and that it's not working), I would trash the containers and start over from the top.
lxc stop <container-name>
lxc rm <container-name>

Before creating the new container (can be same name as you used before if you prefer providing you remove it), I would make sure the VPS is up-to-date sudo apt-get update && sudo apt-get upgrade -y. And once the new container is created, I would do that same command again inside the new container.

[timconsidine]

@privsec said in SMS desktop app:

@timconsidine are there tried and true steps you do to setup certbot with ngoni every time?

I never have a problem with certbot except when I try to create a new certificate too soon after setting up the DNS, i.e. before it is propagated. To restrain my impatience :

• I usually ping -c 5 sub.domain.tld until I get responses.
• I always do nginx -t and systemctl reload nginx to ensure I get up-to-date nginx configs without any errors. Always have to resolve nginx errors before attempting a clean certbot certificate addition.

If you are having nginx or certbot issues, I would make sure system is up to date with sudo apt-get update && sudo apt-get upgrade -y

Oh, and be sure that all nginx and certbot work is done on the VPS, not while you are in the container.

Feel free to ping me in chat if you wish to step through.

[privsec]

@timconsidine said in SMS desktop app:

sudo apt-get update && sudo apt-get upgrade -y

I have wiped the VPS and started over.
But both had the ubuntu 20.04 version

[privsec]

@timconsidine

Ok, I got this working by following your guide, then for lets encrypt, I followed https://haydenjames.io/how-to-set-up-an-nginx-certbot/up to step 4

It is now live!

[timconsidine]

@privsec cool, well done !

[privsec]

@timconsidine This also handles calls just fine

[timconsidine]

@privsec : hadn't tested yet, thanks for doing that !

[robi]

@privsec said in SMS desktop app:

@timconsidine This also handles calls just fine

that would be great for GVoice

[privsec]

@robi Currently, the app only allows for Telnyx or a Twilio API key.

Does GVoice use an API key?

[privsec]

So I was doing some thinking on this, and couldn’t one set up several cron jobs to do the following weekly

• kill all node
• merge the currently cloned repo with the new repo (updates, fixes, bugs)
• restart app

[privsec]

@timconsidine

Are you getting any error messages when you attempt to update the app with the github repo?

When I do, I am getting an error message about node dependencies that I havent gotten before.

I have tried various PPAs of NodeJS and NPM and I still get

E: Unable to correct problems, you have held broken packages.

Any ideas on this one?

[timconsidine]

@privsec no haven't seen that, I will try and see what happens

[privsec]

@timconsidine I am thinking there is something else wrong because even when using the website provided by operation privacy, I cant use the service.

[privsec]

@timconsidine

Out of curiosity, what node version does your lxc use? Mine defaults to 10.x.x

[timconsidine]

@privsec sorry I missed this question
Yes, mine is running 10.x.x

[timconsidine]

@timconsidine in case anyone is interested, updating the app in the LXC container is fairly straight-forward.
I held off for a long time, but eventually bit the bullet.

The steps I took are a bit manual / kludge but it was fastest way.
This assumes you are in root directory of container/installation

1. Stopped the app running
   ps aux | grep node
kill -9 xxxx (process number of running app)

2. "backed up" the original installation by mv VoIP/ VoIP-old/

3. Ran git clone https://github.com/0perationPrivacy/VoIP.git to create a new directory with latest version (0.89)

4. Copied across settings from original deployment
   cp VoIP-old/.env VoIP/.env

5. Edited VoIP/.env to add a new line at the end
   HTTPS = false

6. Started app
   cd VoIP/
node app.js &

This kept all settings (database, domain) and also all data (contacts, messages).

Not extensively tested but it seems to be running ok.