In this lesson, you will learn how hold timers influence the stability of your network by balancing rapid failure detection against the risk of unnecessary link instability.
Understanding Convergence
Convergence is the time required for all routers in a network to agree on the current network topology after a change (like a link failure or a new route advertisement). In Linux, when running protocols like OSPF or BGP via FRR, the speed of this process is heavily gated by hold timers (often referred to as Dead Intervals or Keepalive timers).
The Role of Hold Timers
A hold timer is the duration a router waits to receive a "hello" or "keepalive" packet from a neighbor before declaring that neighbor unreachable.
- Aggressive Timers: Setting a very low hold timer (e.g., 1–3 seconds) allows the kernel to trigger route recalculation almost immediately upon a failure. This minimizes downtime, which is vital for real-time traffic.
- Conservative Timers: Setting a high hold timer (e.g., 30–90 seconds) ignores minor, transient network glitches. This prevents "route flapping"—where the kernel constantly updates its routing table due to intermittent packet loss, which can cause significant CPU spikes and instability.
Configuration Impact
In your FRR configuration, you manage these timers to control your convergence profile. For instance, in OSPF, adjusting the dead-interval directly affects the kernel's ability to pull routes from the Routing Information Base (RIB):
# Example FRR OSPF interface configuration
interface eth0
ip ospf hello-interval 1
ip ospf dead-interval 4
If the link flaps frequently, a dead interval of 4 seconds might cause the kernel to delete and re-insert the route continuously. While this leads to "fast convergence," it also causes "micro-outages" where the kernel routing stack is perpetually recalculating paths rather than forwarding traffic.
If your network infrastructure is prone to minor jitter, an overly aggressive hold timer will decrease network availability rather than improve it. You are essentially choosing between immediate recovery from hard failures and resilience against transient noise. Always profile your physical link reliability before shrinking these timers to sub-second values.
Key Takeaway: Hold timers represent a fundamental trade-off where tighter limits speed up convergence for critical failures but increase the risk of network instability due to transient link jitter.