Serval should be unhealthy if there are no workers on the ClearML queue, such as if a server is failed. We also need a test to make sure this continues to work. Right now (at this moment) there are no workers available for the CPU or the GPU queue for QA (lambert_24gb) - but everything is still saying it's ok.
Serval should be unhealthy if there are no workers on the ClearML queue, such as if a server is failed. We also need a test to make sure this continues to work. Right now (at this moment) there are no workers available for the CPU or the GPU queue for QA (lambert_24gb) - but everything is still saying it's ok.