I understand the automatic detection of issues and automatic swarming of elements is not available yet.
But let's say that all prerequisites are met and swarming is enabled on all DMA's in a cluster.
In case e.g. a DMA goes down unexpectedly or is in a non-healthy state (continuous restarts, ...) and some critical elements would need to be up and running asap, can you then successfully initiate the swarming of these elements to another DMA?
In other words, is there any dependency on the status of the hosting or source DMA to successfully swarm an element?
Hi Koen,
A swarming request to the DMS always goes to the new target host.
This DMA will then in turn try its best to stop/unload the element on the source DMA. However, after some grace period, the new host will get the element up, even if the source DMA is not responding.
Thanks Robbe!
Additional question related to the UI. Currently if a DMA is down, the elements hosted on that DMA are typically missing in the surveyor or in an error state referring to the unavailability of the hosting DMA. I assume that this is not applicable then for the UI where you can swarm an element (Cube). So all elements in that UI will be available for swarming despite the status of the DMA. Correct?