Trees | Indices | Help |
|
---|
|
The interface to the cluster configuration.
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
set |
|
||
string |
|
||
|
|||
list |
|
||
list |
|
||
list |
|
||
list |
|
||
list |
|
||
|
|||
|
|||
|
|||
|
|||
|
|||
(dict, list) |
|
||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
string |
|
||
|
|||
|
|||
string |
|
||
objects.NodeGroup or None |
|
||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
objects.Instance |
|
||
dict |
|
||
|
|||
|
|||
|
|||
objects.Node |
|
||
objects.Node |
|
||
(list, list) |
|
||
list |
|
||
|
|||
|
|||
|
|||
|
|||
dict |
|
||
tuple |
|
||
tuple |
|
||
list |
|
||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
dict |
|
||
|
|||
|
|||
|
|||
|
|||
|
|||
|
|||
objects.Cluster |
|
||
|
|||
|
|||
|
|
|||
|
|
|||
_all_rms a list of all temporary reservation managers |
|||
_temporary_lvs reservation manager for temporary LVs |
|
Generate a MAC for an instance. This should check the current instances for duplicates.
|
Reserve a MAC for an instance. This only checks instances managed by this cluster, it does not check for potential collisions elsewhere.
|
Reserve an VG/LV pair for an instance.
|
Generate a DRBD secret. This checks the current disks for duplicates.
|
Compute the list of all UUIDs and names we have.
|
Generate an unique UUID. This checks the current node, instances and disk names for duplicates.
|
Generate an unique ID. This is just a wrapper over the unlocked version.
|
Return all MACs present in the config.
|
Return all DRBD secrets present in the config.
|
Compute duplicate disk IDs
|
Verify function.
|
Verify function. This is just a wrapper over _UnlockedVerifyConfig.
|
Convert the unique ID to the ID needed on the target nodes. This is used only for drbd, which needs ip/port configuration. The routine descends down and updates its children also, because this helps when the only the top device is passed to the remote node. This function is for internal use, when the config lock is already held. |
Convert the unique ID to the ID needed on the target nodes. This is used only for drbd, which needs ip/port configuration. The routine descends down and updates its children also, because this helps when the only the top device is passed to the remote node.
|
Adds a new port to the available port pool.
|
Returns a copy of the current port list.
|
Allocate a port. The port will be taken from the available port pool or from the default port range (and in this case we increase highest_used_port).
|
Compute the used DRBD minor/nodes.
|
Compute the used DRBD minor/nodes. This is just a wrapper over _UnlockedComputeDRBDMap.
|
Allocate a drbd minor. The free minor will be automatically computed from the existing devices. A node can be given multiple times in order to allocate multiple minors. The result is the list of minors, in the same order as the passed nodes.
|
Release temporary drbd minors allocated for a given instance.
|
Release temporary drbd minors allocated for a given instance. This should be called on the error paths, on the success paths it's automatically called by the ConfigWriter add and update functions. This function is just a wrapper over _UnlockedReleaseDRBDMinors.
|
Get the configuration version.
|
Get cluster name.
|
Get the hostname of the master node for this cluster.
|
Get the IP of the master node for this cluster.
|
Get the master network device for this cluster.
|
Get the file storage dir for this cluster.
|
Get the hypervisor type for this cluster.
|
Return the rsa hostkey from the config.
|
Get the default instance allocator for this cluster.
|
Get cluster primary ip family.
|
Lookup a node group's UUID.
|
Lookup a node group.
|
Get the configuration of all node groups.
|
Get a list of node groups.
|
Add an instance to the config. This should be used after creating a new instance.
|
Ensures a given object has a valid UUID.
|
Mark the instance status to up in the config.
|
Remove the instance from the configuration.
|
Rename an instance. This needs to be done in ConfigWriter and not by RemoveInstance combined with AddInstance as only we can guarantee an atomic rename.
|
Mark the status of an instance to down in the configuration.
|
Get the list of instances. This function is for internal use, when the config lock is already held. |
Get the list of instances.
|
Attempt to expand an incomplete instance name.
|
Returns information about an instance. This function is for internal use, when the config lock is already held. |
Returns information about an instance. It takes the information from the configuration file. Other information of an instance are taken from the live systems.
|
Get the configuration of all instances.
|
Add a node to the configuration.
|
Remove a node from the configuration.
|
Attempt to expand an incomplete instance name.
|
Get the configuration of a node, as stored in the config. This function is for internal use, when the config lock is already held.
|
Get the configuration of a node, as stored in the config. This is just a locked wrapper over _UnlockedGetNodeInfo.
|
Get the instances of a node, as stored in the config.
|
Return the list of nodes which are in the configuration. This function is for internal use, when the config lock is already held.
|
Return the list of nodes which are in the configuration.
|
Return the list of nodes which are online.
|
Return the list of nodes which are not vm capable.
|
Get the configuration of all nodes.
|
Get the number of current and maximum desired and possible candidates.
|
Get the number of current and maximum possible candidates. This is just a wrapper over _UnlockedGetMasterCandidateStats.
|
Try to grow the candidate pool to the desired size.
|
Run upgrade steps that cannot be done purely in the objects. This is because some data elements need uniqueness across the whole configuration, etc. Warning: this function will call _WriteConfig(), but also DropECReservations so it needs to be called only from a "safe" place (the constructor). If one wanted to call it with the lock held, a DropECReservationUnlocked would need to be created first, to avoid causing deadlock. |
Distribute the configuration to the other nodes. Currently, this only copies the configuration file. In the future, it could be used to encapsulate the 2/3-phase update mechanism. |
Return the values needed by ssconf.
|
Wrapper using lock around _UnlockedGetSsconf().
|
Return the volume group name.
|
Set the volume group name.
|
Return DRBD usermode helper.
|
Set DRBD usermode helper.
|
Return the mac prefix.
|
Returns information about the cluster
|
Check if in there is at disk of the given type in the configuration.
|
Notify function to be called after updates. This function must be called when an object (as returned by GetInstanceInfo, GetNodeInfo, GetCluster) has been updated and the caller wants the modifications saved to the backing store. Note that all modified objects will be saved, but the target argument is the one the caller wants to ensure that it's saved.
|
Drop per-execution-context reservations
|
Trees | Indices | Help |
|
---|
Generated by Epydoc 3.0.1 on Mon Dec 20 13:26:50 2010 | http://epydoc.sourceforge.net |