Hi,

I have two dedicated nodes for slurm, node1 and node2.

I have created the following.

*Role*

        

*SlurmCTLD*

        

*SlurmDBD*

        

*Mariadb Server for accounting storage*

*Primary*

        

Node1

        

Node2

        

Node2

*Backup*

        

Node2

        

Node1

        

-

Shared NFS Storage from an NFS Server, for StateSaveLocation.

I want to know what if Node2 goes down. I have read in the documentation that if slurmdbd does down, slurmctld can still hold back the accounting info and when the slurmdbd is back up, it will get it passed on and written to the backend database ( not the exact words, but in that vein).

Just want to know what if node2 goes down and the backup slurmdbd in node1 takes over. Will it fail instantaneously or keep logging the data in it's memory and write back to the DB when it is back up ?

Hope I could explain what I mean.

Thanks & regards,

Richard.

Reply via email to