Initial Node UI: Starting with the terminal

[Southside]

Add works OK, Start Kill and Remove not so much Quit works fine.
Got some messages over writing the table but that cleans up by running the arrow up and down.

I installed this alacritty that I am not used to, I need to remove that entirely and see how it goes with the std Ubuntu terminal

UPDATE:

Add and Remove work fine, this cant connect to a network at the moment so obv Start and Kill cant be expected to work. Looking good so far

works from the command line as well

need to suppress stdout though

apart from that it looks good.
when can we test it for real?

[happybeing]

@joshuef

@happybeing looks like the linux terminal start setup is off there.

Can you elaborate? It’s a new Ub24 install but I copied user settings over from my previous Ub22 system.

EDIT: @joshuef my system won’t boot this morning. Anything the TUI messes with that I should look at? I should never have given it sudo EDIT: see end

@chriso might the TUI have messed with /var/safenode-manager and if so is it safe/worthwhile to try removing it? I’m looking at what system files have been changed in the past 24 hours, but those were probably changed by running snm itself. Any other clues? EDIT: see…

EDIT: no worries, I think best to reinstall so I can get on.

[roland]

@Southside

Add and Remove work fine, this cant connect to a network at the moment so obv Start and Kill cant be expected to work. Looking good so far

Hey @Southside, what does the start do? Nothing? It is actually plugged in and it should work, though I’ve seen some cases earlier where it was not working, but it should’ve been fixed.

@happybeing

@chriso might the TUI have messed with /var/safenode-manager and if so is it safe/worthwhile to try removing it? I’m looking at what system files have been changed in the past 24 hours, but those were probably changed by running snm itself. Any other clues? EDIT: see…

The TUI uses the node-manager APIs under the hood and it should just access /var/log/safenode + /var/safenode-manager and either one of these dirs

    // Linux:   /home/alice/.config/node-launchpad
    // Windows: C:\Users\Alice\AppData\Roaming\maidsafe\node-launchpad
    // macOS:   /Users/Alice/Library/Application Support/com.maidsafe.node-launchpad

The sudo access is just for creating the services (which should hopefully go away soon)

[joshuef]

@happybeing

EDIT: @joshuef my system won’t boot this morning. Anything the TUI messes with that I should look at? I should never have given it sudo EDIT: see end

@chriso might the TUI have messed with /var/safenode-manager and if so is it safe/worthwhile to try removing it? I’m looking at what system files have been changed in the past 24 hours, but those were probably changed by running snm itself. Any other clues? EDIT: see…

the tui is just a wrapper around the safenode manager, so it’s not doing anything that’s not honestly.

@happybeing

Can you elaborate? It’s a new Ub24 install but I copied user settings over from my previous Ub22 system.

I mean in the code, there’s some detection of terminal things that may not be working well. Would be good to know if it works with another terminal for you.

[happybeing]

@joshuef

I mean in the code, there’s some detection of terminal things that may not be working well. Would be good to know if it works with another terminal for you.

Sorry, but I won’t be touching it again for a while. Maybe it wasn’t what borked my system but I reboot every day and that’s the only unusual thing I did yesterday, and it did mess itself up.

[Southside]

@roland

Hey @Southside, what does the start do? Nothing? It is actually plugged in and it should work, though I’ve seen some cases earlier where it was not working, but it should’ve been fixed.

I picked a node at random and hit “S” – nothing

So I looked at it harder

willie@gagarin:~$ sudo systemctl status safenode7
× safenode7.service - safenode7
     Loaded: loaded (/etc/systemd/system/safenode7.service; enabled; vendor preset: enabled)
     Active: failed (Result: exit-code) since Wed 2024-05-08 14:46:27 BST; 52s ago
    Process: 74421 ExecStart=/var/safenode-manager/services/safenode7/safenode --rpc 127.0.0.1:39841 --root-dir /var/safenode-manager/>
   Main PID: 74421 (code=exited, status=2)
        CPU: 3ms

May 08 14:46:27 gagarin systemd[1]: safenode7.service: Scheduled restart job, restart counter is at 5.
May 08 14:46:27 gagarin systemd[1]: Stopped safenode7.
May 08 14:46:27 gagarin systemd[1]: safenode7.service: Start request repeated too quickly.
May 08 14:46:27 gagarin systemd[1]: safenode7.service: Failed with result 'exit-code'.
May 08 14:46:27 gagarin systemd[1]: Failed to start safenode7.

I get this on all my nodes now - cant get any started on this box that i tried out launchpad on. I started 20 nodes on another box here - using safenode-manager and they are working correctly if somewhat feebly AFAICS

EDIT: I have no ~/.config/node-launchpad on my main machine.
Thats yet another SAFE/Autonomi directory to keep an eye on - along with ~/.local/bin , /var/safenode-manager, /var/log/safenode, ~/.local/share/safe/client - did I miss any? Lots of places to check for something no right…

I know proper documentation will come along soon but it would have been nice to know about the supposed prescence of ~/.config/node-launchpad when Im trying to make this work.

:CLARIFICATION
The node 7 shown above is a set of 10 I tried to start manually last night after I remembered

@maidsafe

What it is not

A viable network! Don’t expect nodes started here to connect to anything at the moment. The main stable network is down while we address some alpha-punch issues. But soon you’ll be able to use this there.

so I didnt expect to actually run nodes with this - didnt stop me trying to start nodes that had already been added by launchpad…

[roland]

@Southside

EDIT: I have no ~/.config/node-launchpad on my main machine.

That should be changed with the latest PR. It should go inside ~/.local/share/safe/. I’ll look into the start issue there.

[joshuef]

@happybeing

Sorry, but I won’t be touching it again for a while. Maybe it wasn’t what borked my system but I reboot every day and that’s the only unusual thing I did yesterday, and it did mess itself up.

Fair enough. It could be the way the service files are being managed there is causing some issues perhaps. We’ll hopefully get to the bottom of it soon enough, and we’ll have sudo requirement removed via this pr, which should make that feel a bit safer at least (and make debugging what could be happening her ea bit easier).

[chriso]

@Southside

EDIT: I have no ~/.config/node-launchpad on my main machine.
Thats yet another SAFE/Autonomi directory to keep an eye on - along with ~/.local/bin , /var/safenode-manager, /var/log/safenode, ~/.local/share/safe/client - did I miss any? Lots of places to check for something no right…

If you can cut us a bit of a break here, there is a necessity to have different directories when it comes to the distinction between system-wide vs. user-mode. Soon, we’ll offer the option to run services in user-mode, but the option for running system-wide will still be there. This is because system-wide still makes sense in some cases. Services in user-mode require an active user session, which is not appropriate for all use cases. Plus, there are some service managers, like OpenRC, that don’t support user-mode services.

I’ll need to speak to Roland about ~/.config/node-launchpad. As the name implies, ~/.config is for configuration, not data, so nothing related to the node manager data should be going there. If anything, this would be for things like custom key-bindings related to the TUI itself.