Study Confluent CCAAK Acks and ISR Safety: key concepts, common traps, and exam decision cues.
On this page
This lesson is the highest-yield durability lane on CCAAK. Many wrong answers sound operationally convenient because they increase availability or lower latency, but they quietly weaken write safety.
Durability chooser
If the question is really about…
Strongest first focus
safest normal write acknowledgement
acks=all plus healthy ISR
how many caught-up replicas are required
min.insync.replicas
resilience to broker loss
replication factor
unsafe availability trade-off
unclean leader election
What the exam is really testing
If the scenario shows…
Strong reading
critical data path
durability settings are under test
write failures during degraded replication
ISR and min ISR interaction matters
request for higher availability at any cost
unsafe durability trade-offs may be the trap
replication factor change only
total copies are not the whole durability story
Common traps
Trap
Better rule
treating replication factor as enough by itself
safe writes still depend on acks and ISR health
raising min ISR without understanding write availability impact
stronger safety can also block writes in degraded states
enabling unclean leader election to hide cluster instability
fast recovery can mean data loss
Decision order that usually wins
Start with the acknowledgement path: if the question is about safest accepted writes, think acks=all first.
Then check whether enough replicas are actually caught up: ISR health determines whether that safety is real.
Use min.insync.replicas as the write gate, not as a generic replica-count setting.
Treat unclean leader election as a last-resort availability trade-off, not as a durability feature.