0.2.0

mark apps that exceed the cluster capacity with a custom pod conditio…

0.1.3

Informer goroutine fatal logging (#28)

* Wraps the scheduler informer and other background goroutines with the fatal logging handler

* Refactors the SparkSchedulerExtender Start() and doStart() methods to be similar in pattern to the rest of the codebase

* Moves the checkDemandCRDExists call to run at least once before we launch the SparkSchedulerExtender doStart loop in its goroutine

0.1.2

create demand resources for fifo failures as well (#25)

0.1.1

Add failure-earlier-drivers as a separate category of failures (#24)

Fixes #16

0.1.0

update spark submit test script (#21)

0.0.4

Exclude already completed pods from being included in the overhead ca…

0.0.3

Debug log for large overhead containers (#17)

* log if high overhead pod detected

* log more

* cpu in milli

* debug log

* format

* reuse var

* better msg

* log param

* move cpu + mem outside

* rm line

* move vars to top

0.0.2

Fix: release reservations for terminated pods (#13)

* release reservations for terminated pods

* Update resource.go

0.0.1

fix publish (#8)

* do not checkout artifacts

* ignore artifacts