FA-071 / Distributed coordination / Open access

Failover routes a session behind its observed version · case 01

The fastest replica returns data older than a version this session has already seen.

Verified by executionVariant 1 · 7 checks per implementationDownload source bundle ↓JSON ↗

ROOT CAUSE

Replica selection ignores the session's minimum acceptable applied version.

THE FAILURE

Replica selection ignores the session's minimum acceptable applied version.

Unsuccessful approach: Always choosing the newest replica preserves monotonicity but violates the routing contract when a faster eligible replica exists.

Case contract

Replicas are [ID,applied version,latency]. Return [ID,version] for the minimum (latency,ID) among versions >= session floor, or None if no replica qualifies. Version order is globally comparable in this model.

Why this case matters

Models session-consistent replica routing after a client moves between read endpoints, including the distinction between a required consistency floor and unnecessarily seeking the newest state.

1 / The failure

Exit 1

"""Failure Map reference implementation. Python standard library only."""
import json

N = 1
observations = []
def solve(floor, replicas):
    eligible = replicas
    return list(min(eligible, key=lambda r: (r[2], r[0]))[:2]) if eligible else None
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
v = N+10
check('stale fast replica', solve(v, [['a', v-1, 1], ['b', v, 10]]), ['b', v])
check('fast enough versus newer', solve(v, [['a', v, N], ['b', v+N, 20]]), ['a', v])
check('all replicas lag', solve(v, [['a', v-1, N]]), None)
check('exact floor accepted', solve(v, [['a', v, N]]), ['a', v])
check('stable latency tie', solve(v, [['b', v+1, N], ['a', v, N]]), ['a', v])
check('empty discovery result', solve(v, []), None)
check('zero-version session', solve(0, [['a', 0, N]]), ['a', 0])
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)

Boundary fixture	Actual	Expected	Outcome
stale fast replica	`['a', 10]`	`['b', 11]`	Failed
fast enough versus newer	`['a', 11]`	`['a', 11]`	Passed
all replicas lag	`['a', 10]`	`None`	Failed
exact floor accepted	`['a', 11]`	`['a', 11]`	Passed
stable latency tie	`['a', 11]`	`['a', 11]`	Passed
empty discovery result	`None`	`None`	Passed
zero-version session	`['a', 0]`	`['a', 0]`	Passed

SHA-256 / cff22615b81cbe823a002b4885533646aa728cb5894e26dde77af25e7b12bae3

2 / The unsuccessful fix

Exit 1

"""Failure Map reference implementation. Python standard library only."""
import json

N = 1
observations = []
def solve(floor, replicas):
    eligible = [r for r in replicas if r[1] >= floor]
    return list(min(eligible, key=lambda r: (-r[1], r[2], r[0]))[:2]) if eligible else None
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
v = N+10
check('stale fast replica', solve(v, [['a', v-1, 1], ['b', v, 10]]), ['b', v])
check('fast enough versus newer', solve(v, [['a', v, N], ['b', v+N, 20]]), ['a', v])
check('all replicas lag', solve(v, [['a', v-1, N]]), None)
check('exact floor accepted', solve(v, [['a', v, N]]), ['a', v])
check('stable latency tie', solve(v, [['b', v+1, N], ['a', v, N]]), ['a', v])
check('empty discovery result', solve(v, []), None)
check('zero-version session', solve(0, [['a', 0, N]]), ['a', 0])
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)

Boundary fixture	Actual	Expected	Outcome
stale fast replica	`['b', 11]`	`['b', 11]`	Passed
fast enough versus newer	`['b', 12]`	`['a', 11]`	Failed
all replicas lag	`None`	`None`	Passed
exact floor accepted	`['a', 11]`	`['a', 11]`	Passed
stable latency tie	`['b', 12]`	`['a', 11]`	Failed
empty discovery result	`None`	`None`	Passed
zero-version session	`['a', 0]`	`['a', 0]`	Passed

SHA-256 / 26bb8d3f828f51faea4aca780a5c84cfd998a1fc11fee198633b48a1aa9149c2

HELD IN THE MEMBER ARCHIVE

The verified repair and its recorded checks are member-only.

This mechanism has 7 recorded checks per implementation. The open-access tier publishes the failure and the unsuccessful fix; the repaired source that passes every check, and the observations that prove it, are available to members.

Every case sharing this mechanism uses the same contract and the same repair, so this one record is held back for all of them.

Member access is invitation-based. Sign in with your invited account to inspect the repair.

Verification & scope

This reproducer isolates one failure mechanism. Results cover the supplied fixtures. Variants within a family share a test contract and should remain grouped when constructing evaluation splits. Related mechanisms with a shared evaluation_group must also remain together; these controlled models are not independent production incidents.

Observations recorded using Python 3.12.14 at 2026-09-29T14:36:50.411210+00:00.

Case digest / 67ca306cf9dadc511c0a577456c5f1f33bb37195272a17584d284a569cb7615f