How Stop Fleet in Activating status?

Hi my fleet every time in Activating status and in Events showing

SERVER_PROCESS_CRASHED
Server Process exited without calling process ending. Exit Code: -1073741515

I found in Debug Fleet Creation Issues, how i understand my problem “Server processes start but fail quickly or report poor health.” And i can`t understand how fix that, how i can stop fleet using describe-runtime-configuration command. Or maybe have another command to stop fleet.

p.s.

On this moment fleet stopped after 01h 41m but its so long time :frowning: Any help will be nice to fix that, stopping fleet Immediately if there is a need.

p.s. 2

Maybe from c++ have some method to say server if something wrong set this fleet to error status.

p.s. 3

Or set some property how match counts server trying restart the fleet.

Hello @Anatoliy -

As you noted, fleet creation will eventually time out / give up if the game server crashes. At that point the fleet will move into an error state. There is currently no method of halting fleet creation in the middle of the process, even if you know in advance it will fail. You simply let the fleet fail.

You can always create another fleet (up to 20) before the first fleet times out if you want to make a bug fix or configuration change.

One thing to note, you will need to make sure your game server process responds to period health requests from GameLift. If not, GameLift will think your game process crashed and fleet provisioning will fail. You can learn more about server-side health checks and how to integrate the GameLift Server SDK here:

Add Amazon GameLift to Your Game Server

Does this help?

Hello MisterGameLift. This answer very help me to more correct understand GameLift. Thank you for help and your time.

Sorry for necroposting. @MisterGameLift are there any updates on the issue? It’s really annoying to wait such period of time

Sorry, unfortunately this has yet been prioritized. As @MisterGameLift said, the developer has the option to create a new fleet while the old one is still ACTIVATING, though I understand this user experience is not ideal. To lessen the chance of the unsuccessful fleet creation, you could:

  • Use GameLift Local to test that your server SDK is properly calling initSDK, processReady, etc. and successfully receiving/responding to health checks.
  • Make sure you create your fleet with TCP port 22 open (you cannot update fleet port settings until the fleet is ACTIVE), so you can ssh into the fleet (see example script), run sudo su, and go to /local/game or /local/whitewater to check the logs to see why your fleet is stuck in ACTIVATING
  • Take a look at this doc to see how to prevent/mitigate other fleet creation errors

@JamesM_Aws Thanks! I suggest adding the second workaround (with pre-opening 22 port) to someplace on this page: Debug GameLift fleet issues - Amazon GameLift Linking this ready to use script also can be quite helpful for someone.

Yesterday was debugging the issue that my binary didn’t find the required binary (libssl and libcrypto) and was built with a newer glibc than available with Amazon Linux 2. Without SSH and ldd it’s almost impossible to debug.

And yet another suggestion - please prepare or publish the link to the Docker image with GameLift local. Now people need to prepare such images manually (even if making such images is really easy work). E.g. I’ve prepared my own public-available image with the latest GameLift local: GitHub - ZaMaZaN4iK/docker-gamelift-local

Thanks for the suggestions. I have recorded them in our backlog.

GLIFT-14863
GLIFT-14864