class CfgUpgrade(object):
Undocumented
Class Method |
|
Undocumented |
Static Method |
|
Undocumented |
Static Method |
|
Ensures hostname matches ssconf value. |
Static Method |
|
Undocumented |
Method | __init__ |
Undocumented |
Method |
|
Replaces disk's dev_type attributes according to the given map. |
Method |
|
Undocumented |
Method |
|
Undocumented |
Method |
|
Re-adds the xen_cmd setting to the configuration. |
Method |
|
Return a conservative value of the exclusive_storage flag. |
Method |
|
Undocumented |
Method |
|
Main program. |
Method |
|
Configures the logging module. |
Method |
|
Undocumented |
Method |
|
Undocumented |
Method |
|
Upgrades the disks' device type. |
Method |
|
Specify the nodes from which a disk is accessible in its definition. |
Method |
|
Undocumented |
Method |
|
Undocumented |
Method |
|
Undocumented |
Method |
|
Undocumented |
Method |
|
Undocumented |
Method |
|
Upgrades the instances' configuration. |
Method |
|
Undocumented |
Method |
|
Undocumented |
Method |
|
Undocumented |
Method |
|
Undocumented |
Method |
|
Upgrades the disks as config top level citizens. |
Method |
|
Undocumented |
Instance Variable | args |
Undocumented |
Instance Variable | config |
Undocumented |
Instance Variable | errors |
Undocumented |
Instance Variable | opts |
Undocumented |
Static Method | _ |
Undocumented |
Static Method | _ |
Undocumented |
Method | _ |
Undocumented |
Method | _ |
Undocumented |
Method | _ |
Recursively compute nodes given a top device. |
Method | _ |
Undocumented |
Method | _ |
Undocumented |
Method | _ |
Undocumented |
Method | _ |
Undocumented |
Replaces disk's dev_type attributes according to the given map.
This can be used for both, up or downgrading the disks.
Return a conservative value of the exclusive_storage flag.
Return True if the cluster or at least a nodegroup have the flag set.
Specify the nodes from which a disk is accessible in its definition.
For every disk that is attached to an instance, get the UUIDs of the nodes that it's accessible from. There are three main cases: 1) Internally mirrored disks (DRBD): These disks are accessible from two nodes, so the nodes list will include these. Their children (data, meta) are also accessible from two nodes, therefore they will inherit the nodes of the parent. 2) Externally mirrored disks (Blockdev, Ext, Gluster, RBD, Shared File): These disks should be accessible from any node in the cluster, therefore the nodes list will be empty. 3) Single-node disks (Plain, File): These disks are accessible from one node only, therefore the nodes list will consist only of the primary instance node.