As we all know right now emhttp is not re-startable once it crashes.  Obviously the first part of the request would be a more resilient webgui.  I understand the cause of most emhttp crashes is often the result of plugins.  But at some point I would consider something being so fragile is no longer the fault of others if you know what I mean.

 

But regardless of that shit happens and it would be very helpful if the webgui would either self-restart (perhaps based on another service polling for life and initiating a restart without a positive response) or at least allow a restart from the console without having to reboot.

Agree ... it would be very nice to be able to do this.  You could do it with v4.7 without issue; but with v5 it causes segfaults.

 

wait, what?!?!  you could on 4.7?  I suppose I never tried, but I thought you couldn't "back then" either [shrug].  Either way, yeah it is not doable now.

 

Truth be told, all the unpredictable GUI crashes I had always occurred when upgrading and stopping/starting Plex ... with the upgraded WebGUI!!!  Even worse the powerdown script wasn't working correctly.

 

Ever since I went back to old-school unraid gui I haven't had any more problems.  But it is always back there, niggling at my brain, just waiting for it to happen the next time I have to stop/start plex

Yes, you can restart emhttp with v4.7.  It's a bit tricky ... I have to look at my notes every time I do it (which has only been 2-3 times) -- but it works fine.    In fact, my main media server is still running 4.7 simply because it's been so rock-solid.  I suppose the next time we have a power failure long enough for the UPS to shut it down I'll upgrade it to v5.0, but that hasn't happened since last December ... my current "Uptime" is 342 days        I've had to restart emhttp twice on that server in the past year; but the process works just fine.    Unfortunately, you can't do that with v5

 

As we all know right now emhttp is not re-startable once it crashes.  Obviously the first part of the request would be a more resilient webgui.  I understand the cause of most emhttp crashes is often the result of plugins.  But at some point I would consider something being so fragile is no longer the fault of others if you know what I mean.

 

But regardless of that shit happens and it would be very helpful if the webgui would either self-restart (perhaps based on another service polling for life and initiating a restart without a positive response) or at least allow a restart from the console without having to reboot.

 

+1

Truth be told, all the unpredictable GUI crashes I had always occurred when upgrading and stopping/starting Plex ... with the upgraded WebGUI!!!  Even worse the powerdown script wasn't working correctly.

 

I know this post is a bit old, but thought I'd add my $0.02 here as I used to have this problem as well.  Don't know if you are still having the issue jumperalex, but this is what fixed it for me.

 

This, at least for me, applies to both the 5.0 rc series w/ SimpleFeatures as well as the 5.0 final series with the new webGUI.  I am using both the Plex plugin and Plex Updater plugin.  I have never had emhttp crash when starting or stopping Plex, but updating Plex to a new version would crash it without fail.  What solved the issue for me was using the Plex plugin to stop Plex, then using the Updater plugin to update, then the Plex plugin again to start Plex.  It works fine and no emhttp crash.  If I leave Plex running and update with the Updater plugin, emhttp will crash 100% of the time.

 

You would think the Updater plugin stops Plex before updating, but I can only assume it doesn't or that something about the way it stops it causes the issue.

Yeah I think I figured that out too via trial and error, and your post lends my credibility to my hunch.  Thanks.