LBaaS API and Object Model improvement¶
https://blueprints.launchpad.net/neutron/+spec/lbaas-api-and-objmodel-improvement
LBaaS needs improved API and object model to provide base line for new API parts that provide advanced use cases such as L7, TLS, HA, etc.
This blueprint describes the changes that should be made to object model so that further design and implementation of L7 switching, TLS and HA feature is possible. The new API that exposes the new entities will be done in an separate extension.
This blueprint is not describing design of L7, TLS API parts but may briefly assume some aspects of possible design of those features. Nor does it describe the changes to the API that will fit with this object model.
Problem Description¶
The “advanced” LB configuration which is supported by all LB vendors, both hw and sw, may consist of multiple service endpoints on one ip address, and multiple pools. Here, ‘service endpoint’ means IP address + port.
The problem with existing API/object model is that it only accounts for single VIP and single pool per loadbalancer.
One of the biggest issue of the existing API and object model is that Pool is playing ‘root object’ role and this solely prevents multiple pools per single configuration.
Proposed Change¶
There are a few things that needs to be done in order to make LBaaS API suitable for addition of TLS, L7 and HA features.
1. This will entirely be a new extension and service plugin. LBaaS V1 code should not be affected, but there may be exceptions.
2. Currently Pool object is the root object(*), which is logically incorrect and confusing. From API perspective that means that creating Pool object will not be a starting point of the workflow.
Member object will add subnet_id as an optional attribute.
4. VIP object will not be used. Its attributes will be added to the LoadBalancer object and the Listener object.
5. LoadBalancer object becomes the root object(*) of the LBaaS object model. It will hold the attributes that pertain to the vip and will be the parent of one or many listeners.
6. Additional object Listener is introduced, which is a placeholder for former VIP parameters such as protocol_port, and protocol, see detailed description below
LoadBalancer and Listener relate as 1:M. Listener will initially only be a child of a LoadBalancer, so its life-cycle is limited to a LoadBalancer. Deleting a LoadBalancer will not be alllowed when it has children Listeners.
In the future, this relationship may become M:N if it is decided it is needed. The M:N change will be an additive change so it will not break contract, however, this blueprint will not attempt this.
7. To prevent entities in the database being out of sync with the backend due to concurrent requests, no operations will be allowed if an entity is in a transient state (PENDING_CREATE, PENDING_UPDATE, PENDING_DELETE).
Pool to Health Monitor relationship will change from M:N to 1:1 for now.
9. Attributes of health monitor and member will stay nearly identical except for references to parents.
10. A new synchronous haproxy driver will be created to easily test out the API and DB changes. It will not be meant for production use. Refactoring the agent haproxy driver will be left to another blueprint.
11. Since there are two types of statuses we should worry about, LoadBalancer and Listener will have both provisioning_status and operating_status fields, while Pool and Member will have an operating_status field.
12. operating_status will be an enum of (‘ONLINE’, ‘OFFLINE’, ‘DEGRADED’, ‘ERROR’)
13. provisioning_status will be an enum of (‘ACTIVE’, ‘PENDING_CREATE’, ‘PENDING_UPDATE’, ‘PENDING_DELETE’, ‘ERROR’)
14. Every update of a LoadBalancer, including updates to and adding/removing children entities, will put the LoadBalancer into a PENDING_UPDATE provisioning_status. No other updates to that LoadBalancer or its children will be allowed until the operation has completed.
(*) Root object is an object that represents ‘service instance’. It could be deployed/undeployed, turned on/off, or capability requirements may be applied to it. It also is a starting point of configuration workflow.
Alternatives¶
https://docs.google.com/a/mirantis.com/document/d/1mTfkkdnPAd4tWOMZAdwHEx7IuFZDULjG9bTmWyXe-zo/edit The documents describes differently structured API that uses ‘single-call’ approach.
https://etherpad.openstack.org/p/neutron-lbaas-api-proposals Brief description of how other APIs for other object models look like
http://lists.openstack.org/pipermail/openstack-dev/2014-November/051147.html Samuel Bercovici’s idea of solving the status issue of shared objects
Data Model Impact¶
LBaaS v1 tables will remain unchanged and the following tables will be added for LBaaS v2 use: * neutron.lbaas_loadbalancers * neutron.lbaas_listeners * neutron.lbaas_pools * neutron.lbaas_members * neutron.lbaas_healthmonitors * neutron.lbaas_sessionpersistences * neutron.lbaas_listener_statistics
lbaas_loadbalancers
id - unique identifier
tenant_id
name
description
vip_port_id - the neutron port tied to the vip (this can be used to get the vip_subnet, and vip_address). This will be stored in the database but not exposed through the API.
vip_subnet_id - the subnet a neutron port should be created on
vip_address - the address of the subnet
provisioning_status
operating_status
admin_state_up
listeners - a list of back-references child listener ids
provider - provider in which this load balancer should be provisioned
lbaas_listeners
id - identity
tenant_id
loadbalancer_id
default_pool_id - ID of default pool. Must have compatible protocol with listener.
protocol - Protocol to load balancer: HTTP, HTTPS, TCP, UDP
protocol_port - port number to listen
admin_state_up - admin state (True or False)
provisioning_status
operating_status
Listener model will later be amended with L7 and TLS-related attributes which are out of scope of this blueprint.
lbaas_pools
id - identity
tenant_id
name
description
protocol - Protocol to load balance
lb_method - load balancing method
healthmonitor_id - id of health monitor
admin_state_up - admin state (True/False). That attribute defines administrative state of the pool on all of the backends where it is actually deployed.
operating_status
lbaas_members
id - identity
tenant_id
address - ip address
pool_id - required parent pool
subnet_id - optional subnet this member is on
protocol_port
weight
operating_status
admin_state_up
lbaas_healthmonitors
id - id
tenant_id
type - (TCP, HTTP)
delay
timeout
max_retries
http_method
url_path
expected_codes
admin_state_up
lbaas_sessionpersistences
pool_id
type - (HTTP_COOKIE, SOURCE_IP, APP_COOKIE)
cookie_name
lbaas_listener_statistics
listener_id
bytes_in
bytes_out
active_connections
total_connections
REST API Impact¶
A separate extension will be created exposing the following resources:
/lbaas/loadbalancers
/lbaas/listeners
/lbaas/pools
/lbaas/pools/{pool_id}/members
/lbaas/healthmonitors
Resource Attributes:
/lbaas/loadbalancers
Attribute Name |
Type |
Access |
Default Value |
Validation Conversion |
Description |
---|---|---|---|---|---|
id |
string (UUID) |
RO, all |
generated |
N/A |
identity |
tenant_id |
string (UUID) |
RW, all (No Update) |
generated |
string |
tenant identity |
name |
string |
RW, all |
‘’ |
string |
Human-readable |
description |
string |
RW, all |
‘’ |
string |
Human-readable |
vip_subnet_id |
string (UUID) |
RW, all (No Update) |
required |
string |
creates vip on this subnet |
vip_address |
string (IP Address) |
RW, all (No Update) |
generated |
IPv4 or IPv6 |
Frontend IP address |
admin_state_up |
bool |
RW, all |
True |
bool |
enabled |
provisioning_status |
string |
RO, all |
N/A |
provisioning status |
|
operating_status |
string |
RO, all |
N/A |
operational status |
Deleting a load balancer will only succeed if it is not a parent of any listeners.
/lbaas/listeners
Attribute Name |
Type |
Access |
Default Value |
Validation Conversion |
Description |
---|---|---|---|---|---|
id |
string (UUID) |
RO, all |
generated |
N/A |
identity |
tenant_id |
string (UUID) |
RW, all |
generated |
string |
tenant identity |
name |
string |
RW, all |
‘’ |
string |
Human-readable |
description |
string |
RW, all |
‘’ |
string |
Human-readable |
loadbalancer_id |
string (UUID) |
RW, all (No Update) |
required |
string |
parent load balancer |
connection_limit |
integer |
RW, all |
-1 |
integer |
max connections to the protocol port |
protocol |
string |
RW, all (No Update) |
required |
‘TCP’,’HTTP’,’HTTPS’ |
listening protocol |
protocol_port |
integer |
RW, all |
required |
0-65535 |
listening port |
admin_state_up |
bool |
RW, all |
True |
bool |
enabled |
provisioning_status |
string |
RO, all |
N/A |
provisioning status |
|
operating_status |
string |
RO, all |
N/A |
operational status |
Note that loadbalancer_id is required for now. This will not preclude later implementations that may want to allow M:N loadbalancer to listeners as this is only an API attribute and thus can be changed from required to optional.
Note that default_pool_id is not specified here as the pool will define its parent listener.
Deleting a Listener will only succeed if it is not the parent of a pool.
/lbaas/pools
Attribute Name |
Type |
Access |
Default Value |
Validation Conversion |
Description |
---|---|---|---|---|---|
id |
string (UUID) |
RO, all |
generated |
N/A |
identity |
tenant_id |
string (UUID) |
RW, all |
generated |
string |
tenant identity |
name |
string |
RW, all |
‘’ |
string |
Human-readable |
description |
string |
RW, all |
‘’ |
string |
Human-readable |
listener_id |
string (UUID) |
RW, all (No Update) |
required |
string |
parent listener |
protocol |
string |
RW, all (No Update) |
required |
‘TCP’,’HTTP’,’HTTPS’ |
protocol to send to members |
lb_algorithm |
string |
RW, all |
required |
‘ROUND_ROBIN’,’LEAST_CONNECTIONS’,SOURCE_IP |
load balancing algorithm |
session_persistence |
dictionary |
RW, all |
{} |
type: ‘SOURCE_IP’,’HTTP_COOKIE’,’APP_COOKIE’, cookie_name: string |
session persistence definition |
admin_state_up |
bool |
RW, all |
True |
bool |
enabled |
operating_status |
string |
RO, all |
generated |
N/A |
operational status |
members |
list |
RO, all |
generated |
N/A |
list of members belonging to this pool |
Note that listener_id is required for now. There will be validation that the listener has only one pool as a child.
This should not preclude a later implementation of M:N listener to pools.
Deleting a pool will not succeed if it is the parent of a health monitor. It will however succeed if it is the parent of any children, and those children will be deleted as well.
/lbaas/pools/{pool_id}/members
Attribute Name |
Type |
Access |
Default Value |
Validation Conversion |
Description |
---|---|---|---|---|---|
id |
string (UUID) |
RO, all |
generated |
N/A |
identity |
tenant_id |
string (UUID) |
RW, all |
generated |
string |
tenant identity |
address |
string (IP) |
RW, all (No Update) |
required |
IPv4 or IPv6 |
IP Address member is listening |
protocol_port |
integer |
RW, all |
required |
0-65535 |
port member is listening |
weight |
integer |
RW, all |
1 |
integer |
traffic distribution weight |
subnet_id |
string (UUID) |
RW, all (No Update) |
None |
string |
subnet to access member port |
admin_state_up |
bool |
RW, all |
True |
bool |
enabled |
operating_status |
string |
RO, all |
generated |
N/A |
operational status |
/lbaas/healthmonitors
Attribute Name |
Type |
Access |
Default Value |
Validation Conversion |
Description |
---|---|---|---|---|---|
id |
string (UUID) |
RO, all |
generated |
N/A |
identity |
tenant_id |
string (UUID) |
RW, all |
generated |
string |
tenant identity |
type |
string |
RW, all (No Update) |
required |
‘HTTP’,’HTTPS’,’PING’,’TCP’ |
type of health check |
pool_id |
string (UUID) |
RW, all (No Update) |
required |
string |
id of pool to monitor |
delay |
integer |
RW, all |
required |
integer |
seconds before health check |
timeout |
integer |
RW, all |
required |
integer |
seconds for a failed check |
max_retries |
integer |
RW, all |
required |
integer |
number of failed checks before member is considered OFFLINE |
http_method |
string |
RW,all |
‘GET’ |
‘GET’,’POST’,’PUT’ |
http verb to send http checks |
url_path |
string |
RW, all |
‘/’ |
string |
url to send http checks |
expected_codes |
string |
RW, all |
‘200’ |
comma delimited HTTP response codes |
expected HTTP response codes for a successful check |
admin_state_up |
bool |
RW, all |
TRUE |
bool |
enabled |
Note that pool_id is a required attribute. Similar to listener_id on the pool object, this does not preclude later implementations of 1:M pool to health monitor relationship.
(*) denotes a required attribute
Security Impact¶
Standard Neutron tenant object ownership rules will apply.
Notifications Impact¶
None
New notifications: - loadbalancer - listener - pool - healthmonitor - member
Other End User Impact¶
Users may have access to a new lbaas resources. The CLI commands will be slightly different depending on which extension that is loaded.
Performance Impact¶
None
IPv6 Impact¶
None
Other Deployer Impact¶
LBaaS V1 and LBaaS V2 can coexist in the codebase, but should not be run at the same time. This will have to be enforced in the code.
No migration path from LBaaS V1 to V2 will be done for this blueprint, however another blueprint should do this as it will be a complicated effort.
Developer Impact¶
A new extension, service plugin, and driver.
Community Impact¶
This change has been in review since Juno. Much discussion has taken place over IRC and the mailing list.
Implementation¶
Assignee(s)¶
- Primary assignees:
brandon-logan
Work Items¶
object model change
new loadbalancer extension for new API
unit tests
Dependencies¶
None
Testing¶
Tempest Tests¶
Functional Tests¶
Functional Tests for the load balancer plugin.
API Tests¶
All new resources added by the extension will have positive and negative tests.
Documentation Impact¶
User Documentation¶
Differences between LBaaS V1 and V2 should be documented.
Developer Documentation¶
There should be a new LBaaS V2 API document section documenting the new resources added by the extension.