Hi
We have a question regarding the RDM datastore and VM migrations. We are using a RDM datastore for some of our VMs.
This is a very nice feature and we can use local disks directly from some of our VMs. The question is how we avoid these VMs to be migrated by the scheduler, we need to keep the vms running in the same hyp in our case.
The problem is when we try any maintenance running:
onehost flush <node>
the scheduler also migrates the VMs which use the RDM datastore, but we do not want that. We are not sure if we have an issue in our DRM setup or if there is another way to lock the VMs (we didn’t execute the onevm lock command in this case).
Our current RDM datastore looks like this:
$ onedatastore show rdm.zorua
DATASTORE 102 INFORMATION
ID : 102
NAME : rdm.zorua
USER : oneadmin
GROUP : oneadmin
CLUSTERS : 101,102
TYPE : IMAGE
DS_MAD : dev
TM_MAD : dev
BASE PATH : /var/lib/one//datastores/102
DISK_TYPE : BLOCK
STATE : READYDATASTORE CAPACITY
TOTAL: : 1M
FREE: : 1M
USED: : 0M
LIMIT: : -PERMISSIONS
OWNER : um-
GROUP : u–
OTHER : —DATASTORE TEMPLATE
ALLOW_ORPHANS=“NO”
CLONE_TARGET=“NONE”
CLONE_TARGET_SHARED=“SELF”
CLONE_TARGET_SSH=“SYSTEM”
DATASTORE_CAPACITY_CHECK=“no”
DISK_TYPE=“BLOCK”
DISK_TYPE_SHARED=“BLOCK”
DISK_TYPE_SSH=“BLOCK”
DS_MAD=“dev”
LN_TARGET=“NONE”
LN_TARGET_SHARED=“NONE”
LN_TARGET_SSH=“SYSTEM”
NAME=“rdm.zorua”
QUATTOR=“1”
RESTRICTED_DIRS=“/”
SAFE_DIRS=“/var/tmp”
TM_MAD=“dev”
TM_MAD_SYSTEM=“ssh,shared”
TYPE=“IMAGE_DS”
This is our default setup.
Thanks in advance for your help!
Álvaro