khopkins
96 Posts
Quote from khopkins on August 29, 2019, 8:27 pm
Hello, we have a 3 node cluster with a pool size of 3 w/min size of 2 and need to swap out the CPU's in the node so it needs to be turned off. What's the best way to handle this? Since we have 2 replica's it should keep things going while the 3rd node is shut down, am I assuming correct? This storage carries live VM data and don't want to loose anything. Running the 3.2.1 version.
Hello, we have a 3 node cluster with a pool size of 3 w/min size of 2 and need to swap out the CPU's in the node so it needs to be turned off. What's the best way to handle this? Since we have 2 replica's it should keep things going while the 3rd node is shut down, am I assuming correct? This storage carries live VM data and don't want to loose anything. Running the 3.2.1 version.
admin
2,930 Posts
Quote from admin on August 29, 2019, 10:16 pm
even if you shut down the node suddenly, there will not be a problem. however it may be a bit better to first move the iSCSI paths away from this node to the other ones before shutdown, this will result in limiting the failure handling to the ceph/osd level rather than also having additional error handling at the iSCSI ha level...but again the system can handle both.
even if you shut down the node suddenly, there will not be a problem. however it may be a bit better to first move the iSCSI paths away from this node to the other ones before shutdown, this will result in limiting the failure handling to the ceph/osd level rather than also having additional error handling at the iSCSI ha level...but again the system can handle both.
khopkins
96 Posts
Quote from khopkins on August 30, 2019, 12:57 pm
Thank you for the information, just wanted to make sure.
Thank you for the information, just wanted to make sure.
