You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi
I have been trying starcluster these days and having some questions regarding starcluster scheduling job:
1: I think the RAM are shared across nodes, right? That means job submitted by starcluster to a specific node can also use RAM from other node when it is insufficient. Am I understand correctly?
2: However, I noticed that ,even some node are totally free, new submitted job goes to node that is being occupied. How comes that happen?
Could you please share some explanation for these. Really appreciated!
The text was updated successfully, but these errors were encountered:
Hi
I have been trying starcluster these days and having some questions regarding starcluster scheduling job:
1: I think the RAM are shared across nodes, right? That means job submitted by starcluster to a specific node can also use RAM from other node when it is insufficient. Am I understand correctly?
2: However, I noticed that ,even some node are totally free, new submitted job goes to node that is being occupied. How comes that happen?
Could you please share some explanation for these. Really appreciated!
The text was updated successfully, but these errors were encountered: