precision-timers

The hold time is three times the interval at which keepalive messages are sent, and the hold time is the maximum number of seconds allowed to elapse between successive keepalive messages that BGP receives from a peer. When establishing a BGP connection with the local routing device, a peer sends an open message, which contains a hold-time value. BGP on the local routing device uses the smaller of either the local hold-time value or the peer’s hold-time value as the hold time for the BGP connection between the two peers.

The default hold-time is 90 seconds, meaning that the default frequency for keepalive messages is 30 seconds. More frequent keepalive messages and shorter hold times might be desirable in large-scale deployments with many active sessions (such as edge or large VPN deployments). To configure the hold time and the frequency of keepalive messages, include the hold-time statement at the [edit protocols bgp] hierarchy level. You can configure the hold time at a logical system, routing instance, global, group, or neighbor level. When you set a hold time value to less than 20 seconds, we recommend that you also configure the BGP precision-timers statement. The precision-timers statement ensures that if scheduler slip messages occur, the routing device continues to send keepalive messages. When the precision-timers statement is included, keepalive message generation is performed in a dedicated kernel thread, which helps to prevent BGP session flaps.

Starting with Junos OS Release 15.2, you can register or unregister keepalives of BGP with the automated keepalive precision timer service of the kernel. This service ensures a reliable generation of keepalives for some configurable maximum period after a switchover of the routing engine from backup to primary until BGP is able to take over the keepalive generation.