Still issue with AMP not showing reason for Minecraft server not restarting

  • 100 Views
  • Last Post 4 weeks ago
  • Topic Is Solved
JonathanL posted this 10 June 2020

As per my previous thread which is NOT solved and the below, no indication of any error on the server, yet AMP still is failing to start the minecraft server with absolutely no indication on the problem may be:

[04:00:10] [Server thread/INFO]: RFTools: server is stopping. Shutting down gracefully
[04:00:11] [Minecraft Debug] : Application state changed from Restarting to Stopped
[04:00:11] [Minecraft Warning] : Server failed to start 5 times, not restarting.
[04:00:11] [Minecraft Debug] : Application state changed from Stopped to Failed
[04:00:11] [Core Debug] : Task Restarting Server... () ended: Finished

Full log here: https://hastebin.com/omaxazahoq.sql

Order By: Standard | Newest | Votes
Mike posted this 10 June 2020

Can you triple check the version number for that specific instance? I'm not seeing the log output I'd expect to see. Posting a full log rather than an excerpt would be helpful.

JonathanL posted this 10 June 2020

v1.9.9.8, built 28/05/2020 10:37

JonathanL posted this 10 June 2020

Full logs:

https://hastebin.com/cepexiyoya.log

Mike posted this 10 June 2020

Can you:

  • Wipe your logs
  • Restart AMP
  • Trigger the issue using a new schedule task

The idea being that everything from AMP starting up to the issue is in a single log, currently it's a rolled-over log.

JonathanL posted this 4 weeks ago

This doesn't seem to have been happening for a while.

The following changed on my side which might have caused the fix:

  • The Minecraft instance is now starting off an SSD
  • We alleviated TPS problems
  • Possibly newer AMP versions installed.

My hunch is that it was the moving to the SSD which made the difference.

While I will close this issue and remove my negative Google review (I'm not one to hold grudges). For your benefit, so that you can avoid other potential issues with people posting bad reviews in the future, I will explain the cause of my frustration that resulted in me writing the negative review in the first place.

I am a professional C# developer, I understand software development and understand that building a product which interacts with 3rd party software is not trivial and some aspects are outside of your control.

The behaviour I was experiencing was:

  • AMP performs scheduled shutdown
  • AMP fails to start up the process again. AMP only logs a line saying it supposedly tried 5 times in less than a second, but there is nothing in the logs, neither the Minecraft or AMP ones, about these supposed 5 attempts or why they failed.
  • This never happened during YEARS of using McMyAdmin with the exact same server, saved world and daily restarts, I literally moved the Minecraft folder as is into the AMP instance.

You claimed the problem was due to errors logged by Minecraft during shutdown, but the errors didn't always happen had always existed when we used McMyAdmin and that never had a problem just starting it up again.

I tried to point out that my concern is that AMP claims to restart the server 5 times before giving up, but there was nothing in any logs anywhere supporting this assertion and I felt that unless AMP logged these supposed 5 failed attempts then there seems to be an issue AMP. But you seemed to only focus on the errors which occured before the restarts and based on the fact were not an issue with McMyAdmin, were highly unlikely to be relevant. You were trying to throw the blame back at me by citing errors on my side and then proceeded to ignore my explanations that the errors are not relevant.

Then you closed my in initial support thread at which made me highly upset because despite patience from my side, you seemed to be disregarding all the information I was providing and seemed to only want to get rid of me. It was then that I put a public facing poor review on your Google business listing and posted this new thread, and only then did you stop treating me as a probably tech unsavvy user despite my having pointed out many times that I am quite tech savvy.

Shortly after we put in the SSD (not even a fancy one, it's an old 120GB SATA one) the problem seemed to stop, the restarts stopped failing. My guess is that the faster disk access side stepped whatever random problem was making AMP incorrectly think that the process startup was failing.

Anyway, use my customer service and tech recommendations here or don't use them, AMP doesn't seem to be misbehaving for me anymore, but your general demeanour at the time towards me left a very poor impression of you upon me.

Mike posted this 4 weeks ago

This is why I asked for further logs to see if there was anything showing up. So far as I can tell the logs you provided previously were roll-over logs that did not contain all of the necessary information.

I've taken a re-read of the previous threads(s), and so far as I can tell the conclusion was that AMP was functioning as intended (but showing the wrong error message, it should have been saying that it's not going to bother trying again because it failed so fast rather than saying that it exhausted the retry limit which appears to be the source of a large amount of the confusion)

So to be absolutely specific, the intended behaviour was/is the following:

  • If the application crashes/stops unexpectedly, AMP will try and restart it.
  • If that first attempt to restart it fails within 5 seconds, AMP will make no further attempts. (It assumes the application is in a completely non-functional state) - there was an issue where it would say "5 attempts made" incorrectly when actually it's decided after a single try that it's going to not take any further action.
  • So long as subsequent retry attempts fail after the 5 second mark (some signs of life), AMP will continue to attempt to restart the application up to 5 times in total.
  • Once the application is in the 'Ready' state, the restart counter is reset.

The thing to remember is that in the overwhelming number of circumstances, the server crashing and then failing to start when retried is down to environmental factors. If we take it as red for example that moving to an SSD or some other change is what alleviated the issue it would still have not been our responsibility to diagnose a failure of the Minecraft server application when AMP is functioning as designed.

We can offer support for "Is AMP functioning as designed and doing what it's supposed to" - but the very shortly after the moment that the issue is with the application within AMP then our support has to stop there because it would absolutely not be feasible to provide support beyond that scope at the price level that we charge. All other external factors are to be considered out-of-scope.

Close