iNotify error when multiple Zilla instances are started in K8s Pods on a Portainer.io host

[vordimous]

Describe the bug
Getting iNotify error when multiple Zilla instances are started in K8s Pods on a Portainer.io host. The issue is likely due to a limitation imposed by the host as described in this Portainer forum post.

java.io.IOException: User limit of inotify instances reached or too many open files
	at java.base/sun.nio.fs.LinuxWatchService.<init>(LinuxWatchService.java:62)
	at java.base/sun.nio.fs.LinuxFileSystem.newWatchService(LinuxFileSystem.java:53)
	at io.aklivity.zilla.runtime.engine@0.9.82/io.aklivity.zilla.runtime.engine.internal.registry.FileWatcherTask.<init>(FileWatcherTask.java:52)
	at io.aklivity.zilla.runtime.engine@0.9.82/io.aklivity.zilla.runtime.engine.Engine.<init>(Engine.java:216)
	at io.aklivity.zilla.runtime.engine@0.9.82/io.aklivity.zilla.runtime.engine.EngineBuilder.build(EngineBuilder.java:147)
	at io.aklivity.zilla.runtime.command.start@0.9.82/io.aklivity.zilla.runtime.command.start.internal.airline.ZillaStartCommand.run(ZillaStartCommand.java:161)
	at io.aklivity.zilla.runtime.command@0.9.82/io.aklivity.zilla.runtime.command.internal.ZillaMain$Invoker.invoke(ZillaMain.java:69)
	at io.aklivity.zilla.runtime.command@0.9.82/io.aklivity.zilla.runtime.command.internal.ZillaMain.invoke(ZillaMain.java:40)
	at io.aklivity.zilla.runtime.command@0.9.82/io.aklivity.zilla.runtime.command.internal.ZillaMain.main(ZillaMain.java:34)

Additional context
The fix is likely to update the host setting, but in the majority of cases when a zilla image is deployed to a production cluster the running zilla instances won't need to watch the filesystem for changes. Any rollout of updates to a zilla config would result in tearing down pods and creating new ones with the new config. Adding a default option inside the container to not watch the filesystem for changes would also prevent this error from happening.