active-probe-params
Syntax
active-probe-params probe-name { settings { burst-size { size; } data-fill { string; } data-size { size; } dscp-code-points { dscp; } enable-sla-export { interval; } forwarding-class { forwarding-class-name; } loss-priority (low | high | medium-high| medium-low) { } per-packet-loss-timeout { interval; } probe-count { count; } probe-interval { interval; } } }
Hierarchy Level
[edit security advance-policy-based-routing]
Description
Specify parameter settings for an active probe.
Application performance is monitored and measured using active probes.
In active probing, custom packets are sent between a spoke device and a hub device on multiple routes to measure RTT, jitter, and packet loss between the book-ended points. You can configure to send active probes periodically on all the active and passive links.
Active probing starts after the configuration is committed. A configured number of samples are collected and used for measuring the SLA. If there is a violation detected for any application, the probe metrics are evaluated to determine the best possible link for that application traffic in order to meet performance requirements as in the SLA.
Consider the example, where you configure the probe count as 1000, probe interval as 10 seconds, and burst size as 100. Burst count is calculated as probe count/burst size (1000/100 = 10). Burst-count is 10. So, probes are sent in sets of 10 bursts each containing 100 packets.
Burst interval is calculated as probe interval/burst-count (10/10 =1). Burst interval is 1 second. So, a burst is sent every 1 second. The active probe is initiated from the spoke device to the hub device on each of the overlay path.
SaaS application types do not support ingress, egress jitter types.
The SaaS probe packets are HTTP head packets, sent between the spoke device and the SaaS server.
Options
probe-name | Active probe identifier. |
burst-size | Number of probes sent as a burst. This value should be less than or equal to probe-count. The burst-size configuration is ignored for SaaS probing and always burst size will be used as one.
|
data-fill string | Data payload for a probe packet. This is a hexadecimal string, which is used the payload for probe. Not supported for SaaS probing. |
data-size size | Size of the data portion. Not supported for SaaS probing. |
dscp-code-points dscp | DiffServ code point (DSCP) bits value. |
enable-sla-export | Time Interval (in seconds) at which the active probe data to be exported to controller. This option is disabled by default.
|
forwarding-class forwarding-class-name | Name of the forwarding class
|
loss-priority level | Map packet values to a loss priority. Loss priority allows you to set the priority for dropping packets. Typically, you mark packets exceeding some service level with a high loss priority—that is, a greater likelihood of being dropped. Level can be one of the following:
|
per-packet-loss-timeout | Time interval between two consecutive SaaS probes.
|
probe-count count | Number of samples required to be collected for an SLA measurement. For SaaS applications, probing is to the actual server, we recommend to use three to four samples per probe window.
|
probe-interval interval | Time interval between successive probes. Starting in Junos OS release 20.4R1, the upper limit of the probe interval is changed to 60 seconds. For SaaS applications, we recommend to configure 60 seconds interval to avoid aggressive probing to an actual SaaS server.
|
Required Privilege Level
services—To view this statement in the configuration.
services-control—To add this statement to the configuration.
Release Information
Statement introduced in Junos OS Release
18.2R1. The options forwarding-class
and loss-priority
are introduced in Junos OS Release 19.2R1.