Nodes¶
Concept¶
A Node is a logical object managed by the Senlin service. A node can be a member of at most one cluster at any time. A node can be an orphan node which means it doesn’t belong to any clusters. Senlin provides APIs and command line supports to manage node’s cluster membership. Please refer to Cluster Membership for details.
A node has a profile_id
property when created that specifies which
Profile to use when creating a physical object that backs the node.
Please refer to Profiles for the creation and management of
profile objects.
Listing Nodes¶
To list nodes that are managed by the Senlin service, you will use the command openstack cluster node list. For example:
$ openstack cluster node list
+----------+--------+-------+--------+------------+-------------+...
| id | name | index | status | cluster_id | physical_id |
+----------+--------+-------+--------+------------+-------------+...
| e1b39a08 | node1 | -1 | ACTIVE | | 89ce0d2b |
| 57962220 | node-3 | -1 | ACTIVE | | 3386e306 |
| b28692a5 | stack1 | 1 | ACTIVE | 2b7e9294 | fdf028a6 |
| 4be10a88 | stack2 | 2 | ACTIVE | 2b7e9294 | 7c87f545 |
+----------+--------+-------+--------+------------+-------------+...
Note that some columns in the output table are short ID of objects. Senlin
command line use short IDs to save real estate on screen so that more useful
information can be shown on a single line. To show the full ID in the list,
you can add the option --full-id
to the command.
Sorting the List¶
You can specify the sorting keys and sorting direction when list nodes,
using the option --sort
. The --sort
option accepts a
string of format key1[:dir1],key2[:dir2],key3[:dir3]
, where the keys used
are node properties and the dirs can be one of asc
and desc
. When
omitted, Senlin sorts a given key using asc
as the default direction.
For example, the following command sorts the nodes using the status
property in descending order:
$ openstack cluster node list --sort status:desc
When sorting the list of nodes, you can use one of index
, name
,
status
, init_at
, created_at
and updated_at
.
Filtering the List¶
You can specify the option --cluster
to list nodes that
are members of a specific cluster. For example:
$ openstack cluster node list --cluster c3
+----------+---------+-------+--------+------------+-------------+...
| id | name | index | status | cluster_id | physical_id |
+----------+---------+-------+--------+------------+-------------+...
| b28692a5 | stack1 | 1 | ACTIVE | 2b7e9294 | fdf028a6 |
| 4be10a88 | stack2 | 2 | ACTIVE | 2b7e9294 | 7c87f545 |
+----------+---------+-------+--------+------------+-------------+...
Besides these two options, you can add the option --filters
to the command openstack cluster node list to
specify keys (node property names) and values you want to filter the list.
The valid keys for filtering are name
and status
. For example, the
command below filters the list by node status ACTIVE
:
$ openstack cluster node list --filters status=ACTIVE
Paginating the List¶
In case you have a large number of nodes, you can limit the number of nodes
returned from Senlin server each time, using the option --limit
. For example:
$ openstack cluster node list --limit 1
Another option you can specify is the ID of a node after which you want to
see the returned list starts. In other words, you don’t want to see those
nodes with IDs that is or come before the one you specify. You can use the
option --marker
for this purpose. For example:
$ openstack cluster node list --marker 765385ed-f480-453a-8601-6fb256f512fc
With option --marker
and option --limit
, you will be able
to control how many node records you will get from each request.
Creating a Node¶
To create a node, you need to specify the ID or name of the profile to be
used. For example, the following example creates a node named test_node
using a profile named pstack
:
$ openstack cluster node create --profile pstack test_node
+---------------+--------------------------------------+
| Property | Value |
+---------------+--------------------------------------+
| cluster_id | |
| created_at | None |
| data | {} |
| dependents | {} |
| details | None |
| domain_id | None |
| id | 1984b5a0-9dd7-4dda-b1e6-e8c1f640598f |
| index | -1 |
| init_at | 2015-07-09T11:41:18 |
| location | None |
| metadata | {} |
| name | test_node |
| physical_id | None |
| profile_id | 9b127538-a675-4271-ab9b-f24f54cfe173 |
| profile_name | pstack |
| project_id | 333acb15a43242f4a609a27cb097a8f2 |
| role | |
| status | INIT |
| status_reason | Initializing |
| updated_at | None |
| user_id | 5e5bf8027826429c96af157f68dc9072 |
+---------------+--------------------------------------+
When processing this request, Senlin engine will verify if the profile value specified is a profile name, a profile ID or the short ID of a profile object. If the profile is not found or multiple profiles found matching the value, you will receive an error message.
Note that the index
property of the new node is -1. This is because we
didn’t specify the owning cluster for the node. To join a node to an existing
cluster, you can either use the openstack cluster member add
command (Cluster Membership) after the node is created, or specify the
owning cluster upon node creation, as shown by the following example:
$ openstack cluster node create --profile pstack --cluster c1 test_node
The command above creates a new node using profile pstack
and makes it a
member of the cluster c1
, specified using the option --cluster
.
When a node becomes a member of a cluster, it will get a value for its
index
property that uniquely identifies itself within the owning cluster.
When the owning cluster is specified, Senlin engine will verify if the cluster specified is referencing a profile that has the same Profile Type as that of the new node. If the profile types don’t match, you will receive an error message from the openstack cluster command.
Another argument that could be useful when creating a new node is the option
--role
. The value could be used by a profile type
implementation to treat nodes differently. For example, the following command
creates a node with a master
role:
$ openstack cluster node create --profile pstack --cluster c1 \
--role master master_node
A profile type implementation may check this role value when operating the physical object that backs the node. It is okay for a profile type implementation to ignore this value.
The last argument you can specify when creating a new node is the option
--metadata
. The value for this option is a list of
key-value pairs separated by a semicolon (‘;
’). These key-value pairs are
attached to the node and can be used for whatever purposes. For example:
$ openstack cluster node create --profile pstack \
--metadata owner=JohnWhite test_node
Showing Details of a Node¶
You can use the name, the ID or the “short ID” of a node to name a node for show. The Senlin API and engine will verify if the identifier you specified can uniquely identify a node. An error message will be returned if there is no node matching the identifier or if more than one node matching it.
An example is shown below:
$ openstack cluster node show test_node
+---------------+--------------------------------------+
| Field | Value |
+---------------+--------------------------------------+
| cluster_id | None |
| created_at | 2015-07-09T11:41:20 |
| data | {} |
| dependents | {} |
| details | {} |
| domain_id | None |
| id | 1984b5a0-9dd7-4dda-b1e6-e8c1f640598f |
| index | -1 |
| init_at | 2015-07-09T11:41:18 |
| location | None |
| metadata | {} |
| name | test_node |
| physical_id | 0e444642-b280-4c88-8be4-76ad0d158dac |
| profile_id | 9b127538-a675-4271-ab9b-f24f54cfe173 |
| profile_name | pstack |
| project_id | 333acb15a43242f4a609a27cb097a8f2 |
| role | None |
| status | ACTIVE |
| status_reason | Creation succeeded |
| updated_at | None |
| user_id | 5e5bf8027826429c96af157f68dc9072 |
+---------------+--------------------------------------+
From the output, you can see the physical_id
of a node (if it has been
successfully created). For different profile types, this value may be the
ID of an object that is of certain type. For example, if the profile type used
is “os.heat.stack
”, this means the Heat stack ID; if the profile type used
is “os.nova.server
”, it gives the Nova server ID.
An useful argument for the command openstack cluster node show is
the option --details
. When specified, you will get the details about
the physical object that backs the node. For example:
$ openstack cluster node show --details test_node
Checking a Node¶
Once a node has been created, you can use the name, the ID or the “short ID” of a node to name a node for check. senlin-engine performs a profile-specific check operation to get the latest status of the physical resource (for example a virtual machine). If the virtual machine status is not ACTIVE, the node will be set to ERROR status. For example:
$ openstack cluster node check node-biQA3BOM
Recovering a Node¶
After a node has been created and running for a period of time, if the node goes into ERROR status, you can use to try to restore the node to ACTIVE status, using the openstack cluster node recover. The restore operation will delete the specified node and recreate it. For example:
$ openstack cluster node recover node-biQA3BOM
Updating a Node¶
Once a node has been created, you can change its properties using the command
openstack cluster node update. For example, to change the name of a
node, you can use the option --name
, as shown by the following
command:
$ openstack cluster node update --name new_node_name old_node_name
Similarly, you can modify the role
property of a node using the option
--role
. For example:
$ openstack cluster node update --role slave master_node
You can change the metadata associated with a node using the option
--metadata
:
$ openstack cluster node update --metadata version=2.1 my_node
Using the openstack cluster node update command, you can change the profile used by a node. The following example updates a node for switching to use a different profile:
$ openstack cluster node update --profile fedora21_server fedora20_server
Suppose the node fedora20_server
is now using a profile of type
os.nova.server
where a Fedora 20 image is used, the command above will
initiate an upgrade to use a new profile with a Fedora 21 image.
Senlin engine will verify whether the new profile has the same profile type
with that of the existing one and whether the new profile has a well-formed
spec
property. If everything is fine, the engine will start profile update
process.
Adopting a Node¶
In Senlin service, we can adopt an existing resource as a node and create a
profile for this node. To adopt a node, you need to specify the resource
physical ID by setting --identity
and resource
profile_type name by setting --type
. For example, the
following example adopts a server with ID
1177c8e8-8472-4e9d-8f15-1d4866b85b8b
as a node named test_adopt_node
:
$ openstack cluster node adopt --identity \
1177c8e8-8472-4e9d-8f15-1d4866b85b8b --type os.nova.server-1.0 \
--name test_adopt_node
+---------------+--------------------------------------+
| Field | Value |
+---------------+--------------------------------------+
| cluster_id | |
| created_at | 2017-08-16T07:52:50Z |
| data | {} |
| dependents | {} |
| details | None |
| domain_id | None |
| id | f88b1d7d-1e25-4362-987c-52f8aea26520 |
| index | -1 |
| init_at | 2017-08-16T07:52:50Z |
| location | None |
| metadata | {} |
| name | test_adopt_node |
| physical_id | 1177c8e8-8472-4e9d-8f15-1d4866b85b8b |
| profile_id | f9e5e3dd-d4f3-44a1-901e-351fa39e5801 |
| profile_name | prof-test_adopt_node |
| project_id | 138cf3f92bb3459da02363db8d53ac30 |
| role | |
| status | ACTIVE |
| status_reason | Node adopted successfully |
| updated_at | None |
| user_id | 67dc524bfb45492496c8ff7ecdedd394 |
+---------------+--------------------------------------+
The --name
is optional, if omitted, Senlin engine will
generate a random name start with node-
for the node.
The option --role
could be used by a profile type
implementation to treat nodes differently. For example, the following command
adopts a server as a node with a master
role:
$ openstack cluster node adopt --identity \
1177c8e8-8472-4e9d-8f15-1d4866b85b8b --type os.nova.server-1.0 \
--name test_adopt_node --role master
The option --metadata
is a list of
key-value pairs separated by a semicolon (‘;
’). These key-value pairs are
attached to the node and can be used for whatever purposes. For example:
$ openstack cluster node adopt --identity \
1177c8e8-8472-4e9d-8f15-1d4866b85b8b --type os.nova.server-1.0 \
--name test_adopt_node --metadata "key1=value1;key2=value2"
Another option --overrides
support user to override
the node profile properties. For example, the following command can adopt a
server as a node and override the network properties in node’s profile:
$ openstack cluster node adopt --identity \
1177c8e8-8472-4e9d-8f15-1d4866b85b8b \
--type os.nova.server-1.0 \
--override '{"networks":[{"network": "public"}]}'
The option --snapshot
is boolean type. If set, senlin
Senlin engine will create a snapshot for the resource before accept the
resource as a node.
Previewing a Node for Adoption¶
A resource can be previewed before getting adopted as a Senlin node using the
openstack cluster node adopt command with option
--preview
. To preview a node, you need to specify the
resource physical ID by setting --identity
and resource
profile_type name by setting --type
. For example:
$ openstack cluster node adopt --preview \
--identity 1177c8e8-8472-4e9d-8f15-1d4866b85b8b \
--type os.nova.server-1.0
+--------------+----------------------------------------------------------------------+
| Field | Value |
+--------------+----------------------------------------------------------------------+
| node_preview | +------------+-----------------------------------------------------+ |
| | | property | value | |
| | +------------+-----------------------------------------------------+ |
| | | properties | { | |
| | | | "name": "test0", | |
| | | | "availability_zone": "nova", | |
| | | | "block_device_mapping_v2": [], | |
| | | | "image": "6232a7b9-8af1-4dce-8eb5-f2988a0e34bc", | |
| | | | "key_name": "oskey", | |
| | | | "auto_disk_config": false, | |
| | | | "flavor": "1", | |
| | | | "metadata": {}, | |
| | | | "networks": [ | |
| | | | { | |
| | | | "network": "private" | |
| | | | } | |
| | | | ], | |
| | | | "security_groups": [ | |
| | | | "default", | |
| | | | "default" | |
| | | | ], | |
| | | | "config_drive": false | |
| | | | } | |
| | | type | os.nova.server | |
| | | version | 1.0 | |
| | +------------+-----------------------------------------------------+ |
+--------------+----------------------------------------------------------------------+
The option --overrides
support user to override the node
profile properties. For example, the following command can adopt a server
as a node and override the network properties in node’s profile:
$ openstack cluster node adopt --preview --identity \
1177c8e8-8472-4e9d-8f15-1d4866b85b8b \
--type os.nova.server-1.0 \
--override '{"networks":[{"network": "public"}]}'
+--------------+----------------------------------------------------------------------+
| Field | Value |
+--------------+----------------------------------------------------------------------+
| node_preview | +------------+-----------------------------------------------------+ |
| | | property | value | |
| | +------------+-----------------------------------------------------+ |
| | | properties | { | |
| | | | "name": "test0", | |
| | | | "availability_zone": "nova", | |
| | | | "block_device_mapping_v2": [], | |
| | | | "image": "6232a7b9-8af1-4dce-8eb5-f2988a0e34bc", | |
| | | | "key_name": "oskey", | |
| | | | "auto_disk_config": false, | |
| | | | "flavor": "1", | |
| | | | "metadata": {}, | |
| | | | "networks": [ | |
| | | | { | |
| | | | "network": "public" | |
| | | | } | |
| | | | ], | |
| | | | "security_groups": [ | |
| | | | "default", | |
| | | | "default" | |
| | | | ], | |
| | | | "config_drive": false | |
| | | | } | |
| | | type | os.nova.server | |
| | | version | 1.0 | |
| | +------------+-----------------------------------------------------+ |
+--------------+----------------------------------------------------------------------+
The option --snapshot
is boolean type. If set, senlin
Senlin engine will create a snapshot for the resource before accept the
resource as a node.
Deleting a Node¶
A node can be deleted using the openstack cluster node delete command, for example:
$ openstack cluster node delete my_node
Note that in this command you can use the name, the ID or the “short ID” to
specify the node you want to delete. If the specified criteria cannot match
any nodes, you will get a ResourceNotFound
exception. If more than one
node matches the criteria, you will get a MultipleChoices
exception.
See Also¶
Below are links to documents related to node management: