Hello Everyone
I seem to struggle when it comes to the Ceph full-mesh cluster created.
I made one using a bond on each node using the fiber NIC's two interfaces of 25Gbps fiber card.
Later I made an additional corosync link, not sure if this picks up if the other is down.
I have created a pool called "VM-Storage".
That pool consists of 2 times 240Gb SSD's on 3 nodes, meaning 6 osd's in a PG of 256.
When the Windows basic VM located at one of the nodes is to be migrated live it takes to long.
I've done this before and then it was only 8 seconds.
The odd thing is that when the VM is not live, it only takes a second to migrate (literally)!
What am I doing wrong here?
Any assistance would be highly appreciated!
Everything seems to run on ceph, all osd are up (see attachment)
I have another topic on this in the network section:
See here
I seem to struggle when it comes to the Ceph full-mesh cluster created.
I made one using a bond on each node using the fiber NIC's two interfaces of 25Gbps fiber card.
Later I made an additional corosync link, not sure if this picks up if the other is down.
I have created a pool called "VM-Storage".
That pool consists of 2 times 240Gb SSD's on 3 nodes, meaning 6 osd's in a PG of 256.
When the Windows basic VM located at one of the nodes is to be migrated live it takes to long.
I've done this before and then it was only 8 seconds.
The odd thing is that when the VM is not live, it only takes a second to migrate (literally)!
What am I doing wrong here?
Any assistance would be highly appreciated!
Everything seems to run on ceph, all osd are up (see attachment)
I have another topic on this in the network section:
See here
