反对figure spillover
A spillover configuration on the appliance consists of a primary virtual server that is configured with a spillover method, a spillover threshold, and a backup virtual server. Backup virtual servers can also be configured for spillover, creating a chain of backup virtual servers.
The spillover method specifies the operational condition on which you want to base your spillover configuration (for example, the number of established connections, bandwidth, or combined health of the server farm). When a new connection arrives, the appliance verifies that the primary virtual server is up and compares the operational condition with the configured spillover threshold. If the threshold is reached, the spillover feature diverts new connections to the first available virtual server in the backup chain. The backup virtual server manages the connections it receives until the load on the primary falls below the threshold.
If you configure spillover persistence, the backup virtual server continues to process the connections it received, even after the load on the primary falls below the threshold. If you configure spillover persistence and a spillover persistence timeout, the backup virtual server processes connections for only the specified period after the load on the primary falls below the threshold.
Note:通常,溢出如果值触发的屁股ociated with the spillover method exceeds the threshold (for example, number of connections). However with the server-health spillover method, spillover is triggered if the health of the server farm falls below the threshold.
You can configure spillover in one of the following ways:
- Specify a predefined spillover method. Four predefined methods are available, and they fulfill common spillover requirements.
- 反对figure policy based spillover. In policy based spillover, you use a Citrix ADC rule to specify the conditions for spillover to occur. Citrix ADC rules give you the flexibility to configure spillover for various operational conditions.
Use policy based spillover if a predefined method does not satisfy your requirements. If you configure both for a primary virtual server, the policy based spillover configuration takes precedence over the predefined method.
First, you create the primary virtual server and the virtual servers that you need for the backup chain. You set up the backup chain by specifying one virtual server as the backup for the primary (that is, you create a secondary virtual server), a virtual server as the backup for the secondary (that is, you create a tertiary virtual server), and so on. Then, you configure spillover by either specifying a predefined spillover method or creating and binding spillover policies.
For instructions for assigning a virtual server as the backup for another virtual server, see反对figuring a Backup Load Balancing Virtual Server.
反对figure a predefined spillover method
Predefined spillover methods fulfill some of the more common spillover requirements. To use one of the predefined spillover methods, you configure spillover parameters on the primary virtual server. To create a chain of backup virtual servers, you also configure spillover parameters on backup virtual servers.
If the backup virtual servers reach their own threshold values, and the service type is TCP, the Citrix ADC appliance sends clients a TCP reset. For service types HTTP, SSL, and RTSP, it diverts new requests to the redirect URL configured for the primary virtual server. A redirect URL can be specified for only HTTP, SSL, and RTSP virtual servers. If a redirect URL is not configured, the Citrix ADC appliance sends clients a TCP reset (if the virtual server is of type TCP) or an HTTP 503 response (if the virtual server is of type HTTP or SSL).
Note: With RTSP virtual servers, the Citrix ADC appliance uses only data connections for spillover. If the backup RTSP virtual server is not available, the requests are redirected to an RTSP URL and an RTSP redirect message is sent to the client.
To configure a predefined spillover method for a virtual server by using the command line interface
At the command prompt, type:
set lb vserver -soMethod -soThreshold -soPersistence ENABLED -soPersistenceTimeout
Example
set lb vserver Vserver-LB-1 -soMethod Connection -soThreshold 1000 -soPersistence enabled -soPersistenceTimeout 2
To configure a predefined spillover method for a virtual server by using the configuration utility
- Navigate toTraffic Management>Load Balancing>Virtual Servers, and open the virtual server.
- In Advanced Settings, clickProtection, and set the spillover parameters.
反对figure policy based spillover
Spillover policies, which are based on rules (expressions), enable you to configure the appliance for a wider range of spillover scenarios. For example, you can configure spillover based on the virtual server’s response time, or based on the number of connections in the virtual server’s surge queue.
To configure policy based spillover, first create a spillover action. You then select the expression that you want to use in the spillover policy, configure the policy, and associate the action with it. Finally, you bind the spillover policy to a load balancing, content switching, or global server load balancing virtual server. You can bind multiple spillover policies to a virtual server, with priority numbers. The appliance evaluates the spillover policies in ascending order of priority numbers and performs the action associated with the last policy to evaluate to TRUE.
A virtual server can also have a backup action. The backup action is performed if the virtual server does not have one or more backup virtual servers, or if all backup virtual servers are DOWN, disabled, or have reached their own spillover limits.
When a spillover policy results in an UNDEF condition (an exception thrown when the result of policy evaluation is undefined), an UNDEF action is performed. The UNDEF action is always ACCEPT. You cannot specify an UNDEF action of your choice.
反对figuring a Spillover Action
A spillover action is performed when the spillover policy with which it is associated evaluates to TRUE. Currently, SPILLOVER is the only supported spillover action.
To configure policy based spillover by using the command line interface
At the command prompt, type the following commands to configure a spillover policy and verify the configuration:
add spillover action -action SPILLOVER show spillover action
Example
add spillover action mySoAction -action SPILLOVER Done
show spillover action mySoAction 1) Name: mySoAction Action: SPILLOVER Done
Selecting an Expression for the Spillover Policy
策略表达式,您可以使用任何虚拟-server based expression that returns a Boolean value. For example, you can use one of the following expressions:
SYS.VSERVER("vserver").RESPTIME.GT() SYS.VSERVER("vserver").STATE.EQ(“”), and SYS.VSERVER("vserver").THROUGHPUT.LT ()
In addition to the existing functions such as RESPTIME, STATE, and THROUGHPUT, you can use the following virtual server based functions that have been introduced with this feature:
Averagesurgecount
Returns the average number of requests in the surge queues of active services. Returns 0 (zero) if there are no active services. Raises an UNDEF condition if used with a content switching or global server load balancing virtual server.
Activeservices
Returns the number of active services. Raises an UNDEF condition if used with a content switching or global server load balancing virtual server.
Activetransactions
Returns the value of the virtual-server-level counter for current active transactions.
is_dynamic_limit_reached
Returns a Boolean TRUE if the number of connections that the virtual server manages is equal to the dynamically calculated threshold. The dynamic threshold is the sum of the maximum client (Max Clients) settings of the bound services that are UP.
You can use a policy expression to implement any of the predefined spillover methods. The following table maps the predefined spillover methods to the expressions you can use to implement them:
Table 1. Converting predefined spillover methods to policy expressions
Predefined spillover method | Corresponding expression |
---|---|
CONNECTION | SYS.VSERVER(“ ”)。CONNECTIONS, used with the GT(int) arithmetic function. |
BANDWIDTH | SYS.VSERVER(“ ”)。THROUGHPUT, used with the GT(int) arithmetic function. |
HEALTH | SYS.VSERVER(“ ”)。HEALTH, used with the LT(int) arithmetic function. |
DYNAMICCONNECTION | SYS.VSERVER(“ ”)。IS_DYNAMIC_LIMIT_REACHEDNote: If you implement policy based spillover by using the IS_DYNAMIC_LIMIT_REACHED function, you must also configure the predefined DYNAMICCONNECTION method for the virtual server, so that the statistics required for spillover to work are collected. |
反对figuring a Spillover Policy
A spillover policy uses a Boolean expression as a rule to specify the conditions that must be met for spillover to occur.
To configure a spillover policy by using the command line interface
At the command prompt, type the following commands to configure a spillover policy and verify the configuration:
add spillover policy -rule -action [-comment ] show spillover policy
Example
> add spillover policy mySoPolicy -rule SYS.VSERVER("v1").RESPTIME.GT(50) -action mySoAction -comment "Triggers spillover when the vserver's response time is greater than 50 ms." Done > show spillover policy mySoPolicy 1) Name: mySoPolicy Rule: "SYS.VSERVER("v1").RESPTIME.GT(50)" Action: mySoAction Hits: 0 ActivePolicy: 0 Comment: "Triggers spillover when the vserver's response time is greater than 50 ms." Done >
Binding a Spillover Policy to a Virtual Server
You can bind a spillover policy to load balancing, content switching, or global server load balancing virtual servers). You can bind multiple policies to a virtual server, with Goto expressions controlling the flow of evaluation.
To bind a spillover policy to a virtual server by using the command line interface
At the command prompt, type the following commands to bind a spillover policy to a load balancing, content switching, or global server load balancing virtual server and verify the configuration:
bind (lb | cs | gslb) vserver -policyName -priority [-gotoPriorityExpression ] show (lb | cs | gslb) vserver
Example
> bind lb vserver vserver1 -policyName mySoPolicy -priority 5 Done > show lb vserver vserver1 vserver1 (2.2.2.12:80) - HTTP Type: ADDRESS . . . 1) Spillover Policy Name: mySoPolicy Priority: 5 GotoPriority Expression: END Flowtype: REQUEST Done >
反对figuring a Backup Action for a Spillover Event
A backup action specifies what to do when the spillover threshold is reached but one or more backup virtual serves are either not configured or are down, disabled, or have reached their own thresholds.
Note: For the predefined spillover methods that are configured directly on the virtual server (as values of the Spillover Method parameter), the backup action is not configurable. By default, the appliance sends clients a TCP reset (if the virtual server is of type TCP) or an HTTP 503 response (if the virtual server is of type HTTP or SSL).
The backup action is configured on the virtual server. You can configure the virtual server to accept requests (after the threshold specified by the policy is reached), redirect clients to a URL, or simply drop requests even before establishing TCP or SSL connections until the number of requests falls below the threshold. Therefore, lesser memory resources are utilized as the connections are reset even before allocating any data structures.
To configure a backup action for spillover by using the CLI
At the command prompt, type the following commands to configure a backup action and verify the configuration:
set lb vserver -soBackupAction show lb vserver
Example:
set lb vserver vs1 -soBackupAction REDIRECT -redirectURL `http://www.mysite.com/maintenance` Done > show lb vserver vs1 vs1 (10.102.29.76:80) - HTTP Type: ADDRESS State: UP . . . Redirect URL: `http://www.mysite.com/maintenance` . . . Done
To configure a backup action for spillover by using the GUI
- Navigate toTraffic Management>Load Balancing>Virtual Servers, and open the virtual server.
- In Advanced Settings, clickProtection, and then specify a spillover backup action.