Problem:
A newly published Model API seemed stuck in a Starting state for a long time. Inspection of logs showed an error within the model listing the following repeatedly
*** no app loaded. GAME OVER ***
Resolution:
There is retry code that repeatedly tries to restart a Model API that encounters failure during launch. Unfortunately the "Starting" forever can actually mean the Model API's code is just failing on repeat.
This behavior is being improved in version 5.1.0, but for more recent versions you can limit the number of retries.
To limit the retry count in 3.6 til 5.1.0 you can add the following to Central Config:
com.cerebro.domino.modelManager.modelContainer.restartCountLimit
This defines how many times a Model API can fail to launch before being descheduled and not restarting further. Something reasonable might be a value of 3
Comments
0 comments
Please sign in to leave a comment.