modelcontextprotocol/conformance
Watch
1
0
Fork
You've already forked conformance
 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

fix(sse-retry): gate retry timing only on the early side (#375)

Browse source 
The retry MUST ("waiting the given number of milliseconds before
attempting to reconnect") is a lower bound; no spec text bounds reconnect
delay from above. The slightly-late WARNING and >2x FAILURE bands gated
environment latency, which CI load inflates - in python-sdk CI the warning
band fired on ~4-5% of runs under the suite's parallel client spawns.

- delete the late bands: FAILURE only when the reconnect arrives earlier
  than retry minus the 50ms tolerance
- demote the could-not-measure path from WARNING to INFO
- negotiate protocolVersion 2025-11-25, the version whose MUST is under
  test (was 2025-03-26; same fix as 4686e06 for server-sse-polling)

Strictly relaxing: no client that passes today can start failing.

Co-authored-by: Paul Carleton <paulcarletonjr@gmail.com>
This commit is contained in:
Max 2026-09-06 20:18:43 +01:00 • committed by GitHub
commit 61d2ccbc55
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 11 additions and 36 deletions
Download patch file Download diff file Expand all files Collapse all files

47
src/scenarios/client/sse-retry.ts
View file

@ -42,10 +42,8 @@ export class SSERetryScenario implements Scenario {
private pendingToolCallId: number | string | null = null;
private getResponseStream: http.ServerResponse | null = null;
// Tolerances for timing validation
// Tolerance for timing validation (early side only; lateness is not gated)
private readonly EARLY_TOLERANCE = 50; // Allow 50ms early for scheduler variance
private readonly LATE_TOLERANCE = 200; // Allow 200ms late for network/event loop
private readonly VERY_LATE_MULTIPLIER = 2; // If >2x retry value, client is likely ignoring it
async start(_ctx: ScenarioContext): Promise<ScenarioUrls> {
return new Promise((resolve, reject) => {
@ -290,7 +288,8 @@ export class SSERetryScenario implements Scenario {
jsonrpc: '2.0',
id: request.id,
result: {
protocolVersion: '2025-03-26',
// The retry MUST under test was introduced in 2025-11-25 (SEP-1699)
protocolVersion: '2025-11-25',
serverInfo: {
name: 'sse-retry-test-server',
version: '1.0.0'
@ -459,39 +458,22 @@ export class SSERetryScenario implements Scenario {
) {
const actualDelay = this.getReconnectionTime - this.toolStreamCloseTime;
const minExpected = this.retryValue - this.EARLY_TOLERANCE;
const maxExpected = this.retryValue + this.LATE_TOLERANCE;
// The retry MUST is a lower bound ("waiting the given number of
// milliseconds before attempting to reconnect"), so only an early
// reconnect fails; lateness is environment latency and is not gated.
const tooEarly = actualDelay < minExpected;
const slightlyLate = actualDelay > maxExpected;
const veryLate =
actualDelay > this.retryValue * this.VERY_LATE_MULTIPLIER;
const withinTolerance = !tooEarly && !slightlyLate;
let status: 'SUCCESS' | 'FAILURE' | 'WARNING' = 'SUCCESS';
let errorMessage: string | undefined;
if (tooEarly) {
// Client reconnected too soon - MUST violation
status = 'FAILURE';
errorMessage = `Client reconnected too early (${actualDelay.toFixed(0)}ms instead of ${this.retryValue}ms). Client MUST respect the retry field and wait the specified time.`;
} else if (veryLate) {
// Client reconnected way too late - likely ignoring retry field entirely
status = 'FAILURE';
errorMessage = `Client reconnected very late (${actualDelay.toFixed(0)}ms instead of ${this.retryValue}ms). Client appears to be ignoring the retry field and using its own backoff strategy.`;
} else if (slightlyLate) {
// Client reconnected slightly late - not a spec violation but suspicious
status = 'WARNING';
errorMessage = `Client reconnected slightly late (${actualDelay.toFixed(0)}ms instead of ${this.retryValue}ms). This is acceptable but may indicate network delays.`;
}
this.checks.push({
id: 'client-sse-retry-timing',
name: 'ClientRespectsRetryField',
description:
'Client MUST respect the retry field, waiting the given number of milliseconds before attempting to reconnect',
status,
status: tooEarly ? 'FAILURE' : 'SUCCESS',
timestamp: new Date().toISOString(),
errorMessage,
errorMessage: tooEarly
? `Client reconnected too early (${actualDelay.toFixed(0)}ms instead of ${this.retryValue}ms). Client MUST respect the retry field and wait the specified time.`
: undefined,
specReferences: [
{
id: 'SEP-1699',
@ -502,14 +484,7 @@ export class SSERetryScenario implements Scenario {
expectedRetryMs: this.retryValue,
actualDelayMs: Math.round(actualDelay),
minAcceptableMs: minExpected,
maxAcceptableMs: maxExpected,
veryLateThresholdMs: this.retryValue * this.VERY_LATE_MULTIPLIER,
earlyToleranceMs: this.EARLY_TOLERANCE,
lateToleranceMs: this.LATE_TOLERANCE,
withinTolerance,
tooEarly,
slightlyLate,
veryLate,
getConnectionCount: this.getConnectionCount
}
});
@ -518,7 +493,7 @@ export class SSERetryScenario implements Scenario {
id: 'client-sse-retry-timing',
name: 'ClientRespectsRetryField',
description: 'Client MUST respect the retry field timing',
status: 'WARNING',
status: 'INFO',
timestamp: new Date().toISOString(),
errorMessage:
'Could not measure timing - tool stream close time or GET reconnection time not recorded',