replica.lag.time.max.ms — Kafka Broker Configuration
If a follower hasn't sent any fetch requests or hasn't consumed up to the leaders log end offset for at least this time, the leader will remove the follower from isr.
Description
If a follower hasn't sent any fetch requests or hasn't consumed up to the leaders log end offset for at least this time, the leader will remove the follower from isr
Default Values by Kafka Version
| Kafka Version | Default Value |
|---|---|
| 0.8.0 | 10000 |
| 0.8.1 | 10000 |
| 0.8.2 | 10000 |
| 0.9.0 | 10000 |
| 0.10.0 | 10000 |
| 0.10.1 | 10000 |
| 0.10.2 | 10000 |
| 0.11.0 | 10000 |
| 1.0 | 10000 |
| 1.1 | 10000 |
| 2.0 | 10000 |
| 2.1 | 10000 |
| 2.2 | 10000 |
| 2.3 | 10000 |
| 2.4 | 10000 |
| 2.5 | 30000 |
| 2.6 | 30000 (30 seconds) |
| 2.7 | 30000 (30 seconds) |
| 2.8 | 30000 (30 seconds) |
| 3.0 | 30000 (30 seconds) |
| 3.1 | 30000 (30 seconds) |
| 3.2 | 30000 (30 seconds) |
| 3.3 | 30000 (30 seconds) |
| 3.4 | 30000 (30 seconds) |
| 3.5 | 30000 (30 seconds) |
| 3.6 | 30000 (30 seconds) |
| 3.7 | 30000 (30 seconds) |
| 3.8 | 30000 (30 seconds) |
| 3.9 | 30000 (30 seconds) |
| 4.0 | 30000 (30 seconds) |
| 4.1 | 30000 (30 seconds) |
| 4.2 | 30000 (30 seconds) |
Tuning Recommendation
| Profile | Recommended | Why |
|---|---|---|
| broker / throughput | 30000 | Keep the default 30s ISR shrink threshold. Lowering it (e.g., 10s) causes spurious ISR shrinks during normal GC pauses or network hiccups, triggering leader elections and momentarily halting produce. 30s is the right balance. |
| broker / durability | 10000 | Reducing ISR shrink threshold to 10s means a follower that falls behind is removed from ISR faster, forcing acks=all producers to wait for a genuinely in-sync replica rather than an outdated one. This tightens the durability guarantee. |
| broker / latency | 10000 | Detects slow followers in 10s instead of 30s, triggering ISR shrink faster so acks=all requests are not blocked waiting on a lagging replica. |
Related Configs
num.replica.fetchers · min.insync.replicas · unclean.leader.election.enable
Manage Kafka configs across all your clusters with Conduktor Console — view, compare, and update configurations in one place.