Re-scan glob path not working

[willief]

'r' to rescan the glob paths doesn't seem to work anymore.

v0.17.2 ubuntu 22.04

EDIT 'r' works as expected IFF you remember to use the -g flag before the path(s) to the logfiles. Problem is if you don't specify the -g flag, vdash still does 99.9% of what it says on the tin.

https://forum.autonomi.community/t/vdash-dashboard-for-autonomi-nodes/32630/248?u=southside

[happybeing]

I had this once yesterday, thought it might be due to trying out Starlink because at some point it worked again.

Very odd. I'm also on Ubuntu 22.04, except in the cloud which is 20.04. Normally I only run vdash in the cloud but yesterday I ran it locally as well so it might have been then.

Let me know if this is intermittent or constant for you. I need to check that again too.

[happybeing]

A hunch: this might be due to vdash not seeing the keypress if the logfile activity slows or pauses.

I didn't check log activity, but when starting my alpha test nodes things didn't happen as normal, so I'm wondering if there was a reduction in logfile activity.

Keypress only works if the logfiles are being updated - this can be verified by killing all your nodes and trying to quit vdash by pressing 'q' which I often do 🤦‍♂️

Maybe this could be fixed using an event timer in the main 'loop'. Cc @RMcTn

[willief]

Its constant, I just did as you suggested above and killed 50 non-earning nodes and then tried to refresh with 'r'. No luck, I'm afraid.

Just for completeness, this was on my local box but Im pretty sure I have seen this home and away.

[happybeing]

Its constant, I just did as you suggested above and killed 50 non-earning nodes and then tried to refresh with 'r'. No luck, I'm afraid.

That's expected because no nodes means no logfile updates means no keypresses working. Once your nodes are static or killed off you're stuck!

[RMcTn]

Keypress only works if the logfiles are being updated - this can be verified by killing all your nodes and trying to quit vdash by pressing 'q' which I often do 🤦‍♂️

I haven't been able to reproduce this on macOS 14.3.1. I can still quit with 'q' as normal. Even rescanning glob files with 'r' picks up newly started nodes after killing all existing nodes.

I'll try it in Ubuntu at some point later

[RMcTn]

I haven't been able to reproduce this on macOS 14.3.1. I can still quit with 'q' as normal. Even rescanning glob files with 'r' picks up newly started nodes after killing all existing nodes.

'r' rescanning was working because I was using -g "/var/log/safenode/*/safenode.log". This worked on Ubuntu as well, although I couldn't reproduce the quit issue after stopping nodes

[happybeing]

I haven't used node-manager at all. I don't see why it would make a difference (compared to ~/.local/share/safe/node/*/logs/safenode.log) but it might be worth checking.

Thanks for looking into this.

[willief]

Using node-manager, the glob would be /var/log/safenode/safenode*/safenode.log

[drirmbda]

Here is another instance of this issue: On Rocky 9.4, (tmux,) bash, vdash 0.17.6, about 30 nodes running: All keys works except for ‘r’ which appears not to do anything. This is independent of whether I have .local symlinked or not. It does not work when I am not using tmux either.

An additional observation: In case of a symlinked .local directory (related to this issue), after pressing ‘r’ then waiting, and pressing ‘space’, most nodes appear INACTIVE except a few that are/were Shunned.

[Voim4]

I have this issue with two machines. One is running Ubuntu 24.04 LTS and the other Ubuntu 22.04.3 LTS.

Even when adding from 3 to 4 nodes, the r does not appear to do anything.

[r43]

Not working for me too. Windows 11, specs at https://github.com/happybeing/vdash/issues/26

[happybeing]

Thanks everyone for your reports. I rarely see this issue and so have been thinking it might be too do with running more nodes than I do.

@Voim4 @r43 Please can you include how many nodes vdash is monitoring when this happens, and whether if you start with only a few nodes, say 5, you still get no response to r, after day adding one or two new ones.

Everyone, please can you also include how you are starting nodes and where the logs are located.

[r43]

I can add nodes after that, started with 3, one by one, they are appearing, let's hope that issue https://github.com/happybeing/vdash/issues/26 will be solved so I could have better monitoring.

[happybeing]

It would help to have all relevant info in the first report each of you make rather than spread around and not clear what is being replied to. Discussion can then be for follow up, but a reply that says "after that" without it being clear what "that" is wastes time and doesn't add useful information.

I'm pointing this out to help you, because if you want things to happen, they will be more likely to happen the easier you make it for myself or anyone else doing them.

If maintaining vdash doesn't remain fun for me, it won't happen unless someone else enjoys it. Currently it is still fun, but my time is limited and since it works fine for me still, I'm focused on things that I believe are more important for now.

I'll probably get back to vdash for some tidying of PRs before long, but this issue isn't something I experience, so it will need to be clarified further unless someone who is experiencing the decides to pick it up.

Another factor is that if Autonomi switch to RPC analytics and vdash follows, this and #26 will become redundant. I don't see that happening very soon but it is a factor in deciding where best to spend my time.

In the mean time, thanks again everyone for your reports and if anyone does want to have a go at an issue themselves, myself and others will I'm sure provide guidance and support. But this stuff takes a lot of time, especially when starting out, so be prepared for that. 🙏