Skip to content

fix(otel): set explicit span status OK on successful operation and attempt spans #583

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
SilanHe merged 4 commits into from
Aug 4, 2026
Merged

fix(otel): set explicit span status OK on successful operation and attempt spans #583

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
c8defb4
feat(otel): set StatusCode.OK on operation and attempt spans on success
hsilan Jul 31, 2026
8f90e58
Merge branch 'main' into feat/otel-setstatus-ok-on-success
SilanHe Jul 31, 2026
ad07ddb
fix(otel): gate operation-span OK on SUCCEEDED status (addresses AI r…
hsilan Jul 31, 2026
f77d306
Merge branch 'main' into feat/otel-setstatus-ok-on-success
SilanHe Jul 31, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 12 additions & 0 deletions otel-plugin/src/main/java/software/amazon/lambda/durable/otel/ExecutionOtelPlugin.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -308,6 +308,12 @@ public void onOperationEnd(OperationEndInfo info) {
if (info.error() != null) {
span.setStatus(StatusCode.ERROR, info.error().getMessage());
span.recordException(info.error());
} else if ("SUCCEEDED".equals(info.status()) || info.status() == null) {
// Only stamp OK on genuine success. onOperationEnd fires for every terminal status, and
// extractErrorFromOperation returns null for CANCELLED (always) and for FAILED/TIMED_OUT/STOPPED
// with no attached error object — those carry a non-null, non-SUCCEEDED status and must stay UNSET.
// A null status is a successful statusless virtual (FLAT CONTEXT) operation, which is OK.
span.setStatus(StatusCode.OK);
Comment thread
SilanHe marked this conversation as resolved.
Comment thread
SilanHe marked this conversation as resolved.
}
endSpan(span, info.endTimestamp());
} else {
Expand Down Expand Up @@ -350,6 +356,10 @@ public void onOperationEnd(OperationEndInfo info) {
if (info.error() != null) {
continuationSpan.setStatus(StatusCode.ERROR, info.error().getMessage());
continuationSpan.recordException(info.error());
} else if ("SUCCEEDED".equals(info.status()) || info.status() == null) {
// See onOperationEnd (this-invocation branch): only genuine success (or a successful statusless
// virtual operation) is OK; error-less non-success statuses stay UNSET.
continuationSpan.setStatus(StatusCode.OK);
}

endSpan(continuationSpan, info.endTimestamp());
Expand Down Expand Up @@ -438,6 +448,8 @@ public void onUserFunctionEnd(UserFunctionEndInfo info) {
if (!info.succeeded() && info.error() != null) {
span.setStatus(StatusCode.ERROR, info.error().getMessage());
span.recordException(info.error());
} else if (info.succeeded()) {
span.setStatus(StatusCode.OK);
}

endSpan(span, info.endTimestamp());
Expand Down
12 changes: 12 additions & 0 deletions otel-plugin/src/main/java/software/amazon/lambda/durable/otel/InvocationOtelPlugin.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -430,6 +430,12 @@ public void onOperationEnd(OperationEndInfo info) {
if (info.error() != null) {
span.setStatus(StatusCode.ERROR, info.error().getMessage());
span.recordException(info.error());
} else if ("SUCCEEDED".equals(info.status()) || info.status() == null) {
// Only stamp OK on genuine success. onOperationEnd fires for every terminal status, and
// extractErrorFromOperation returns null for CANCELLED (always) and for FAILED/TIMED_OUT/STOPPED
// with no attached error object — those carry a non-null, non-SUCCEEDED status and must stay UNSET.
// A null status is a successful statusless virtual (FLAT CONTEXT) operation, which is OK.
span.setStatus(StatusCode.OK);
Comment thread
SilanHe marked this conversation as resolved.
}
span.end();
Comment thread
SilanHe marked this conversation as resolved.
} else {
Expand Down Expand Up @@ -473,6 +479,10 @@ public void onOperationEnd(OperationEndInfo info) {
if (info.error() != null) {
continuationSpan.setStatus(StatusCode.ERROR, info.error().getMessage());
continuationSpan.recordException(info.error());
} else if ("SUCCEEDED".equals(info.status()) || info.status() == null) {
// See onOperationEnd (this-invocation branch): only genuine success (or a successful statusless
// virtual operation) is OK; error-less non-success statuses stay UNSET.
continuationSpan.setStatus(StatusCode.OK);
}

continuationSpan.end();
Expand Down Expand Up @@ -568,6 +578,8 @@ public void onUserFunctionEnd(UserFunctionEndInfo info) {
if (!info.succeeded() && info.error() != null) {
span.setStatus(StatusCode.ERROR, info.error().getMessage());
span.recordException(info.error());
} else if (info.succeeded()) {
span.setStatus(StatusCode.OK);
}

if (info.endTimestamp() != null) {
Expand Down
74 changes: 74 additions & 0 deletions otel-plugin/src/test/java/software/amazon/lambda/durable/otel/ExecutionOtelPluginTest.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -344,6 +344,80 @@ void userFunctionFailure_setsErrorOnAttemptSpan() {
assertEquals(StatusCode.ERROR, attemptSpan.getStatus().getStatusCode());
}

@Test
void userFunctionSuccess_setsOkOnAttemptSpan() {
plugin.onInvocationStart(new InvocationInfo("req-1", ARN, true, Instant.now()));
plugin.onUserFunctionStart(
new UserFunctionStartInfo("op-1", "compute", "STEP", "Step", null, Instant.now(), false, 1));
plugin.onUserFunctionEnd(new UserFunctionEndInfo(
"op-1", "compute", "STEP", "Step", null, Instant.now(), Instant.now(), false, 1, true, null));
plugin.onInvocationEnd(new InvocationEndInfo("req-1", ARN, true, InvocationStatus.SUCCEEDED, null));

var attemptSpan = spanExporter.getFinishedSpanItems().stream()
.filter(s -> s.getName().contains("attempt"))
.findFirst()
.orElseThrow();
assertEquals(StatusCode.OK, attemptSpan.getStatus().getStatusCode());
}

@Test
void operationSuccess_setsOkOnOperationSpan() {
plugin.onInvocationStart(new InvocationInfo("req-1", ARN, true, Instant.now()));
plugin.onOperationStart(new OperationInfo("op-1", "step-ok", "STEP", "Step", null, Instant.now(), null, false));
plugin.onOperationEnd(new OperationEndInfo(
"op-1", "step-ok", "STEP", "Step", null, Instant.now(), Instant.now(), "SUCCEEDED", null, false, null));
plugin.onInvocationEnd(new InvocationEndInfo("req-1", ARN, true, InvocationStatus.SUCCEEDED, null));

var operationSpan = spanByName(spanExporter.getFinishedSpanItems(), "step-ok");
assertEquals(StatusCode.OK, operationSpan.getStatus().getStatusCode());
}

@Test
void operationEnd_withNonSuccessStatusAndNoError_leavesOperationSpanUnset() {
// onOperationEnd fires for every terminal status. A CANCELLED operation (or an error-less
// FAILED/TIMED_OUT/STOPPED) carries a non-null, non-SUCCEEDED status with a null error. It must NOT be
// stamped OK — the span status stays UNSET.
plugin.onInvocationStart(new InvocationInfo("req-1", ARN, true, Instant.now()));
plugin.onOperationStart(
new OperationInfo("op-cancel", "step-cancel", "STEP", "Step", null, Instant.now(), null, false));
plugin.onOperationEnd(new OperationEndInfo(
"op-cancel", "step-cancel", "STEP", "Step", null, Instant.now(), Instant.now(), "CANCELLED", null,
false, null));
plugin.onInvocationEnd(new InvocationEndInfo("req-1", ARN, true, InvocationStatus.SUCCEEDED, null));

var operationSpan = spanByName(spanExporter.getFinishedSpanItems(), "step-cancel");
assertEquals(StatusCode.UNSET, operationSpan.getStatus().getStatusCode());
}

@Test
void operationEnd_withoutStart_nonSuccessStatusAndNoError_leavesContinuationSpanUnset() {
// Same guard on the continuation-span branch (operation completed between invocations): an error-less
// TIMED_OUT terminal status must NOT be stamped OK.
plugin.onInvocationStart(new InvocationInfo("req-2", ARN, false, Instant.now()));
plugin.onOperationEnd(new OperationEndInfo(
"op-cb-timeout", "my-callback", "CALLBACK", "Callback", null, Instant.now(), Instant.now(),
"TIMED_OUT", null, false, null));
plugin.onInvocationEnd(new InvocationEndInfo("req-2", ARN, false, InvocationStatus.SUCCEEDED, null));

var continuationSpan = spanByName(spanExporter.getFinishedSpanItems(), "my-callback");
assertEquals(StatusCode.UNSET, continuationSpan.getStatus().getStatusCode());
}

@Test
void operationEnd_withNullStatusAndNoError_setsOkOnOperationSpan() {
// A successful statusless virtual (FLAT CONTEXT) operation fires onOperationEnd with a null operation ->
// null status and null error. This is genuine success and must be stamped OK.
plugin.onInvocationStart(new InvocationInfo("req-1", ARN, true, Instant.now()));
plugin.onOperationStart(
new OperationInfo("op-ctx", "my-ctx", "CONTEXT", null, null, Instant.now(), null, false));
plugin.onOperationEnd(new OperationEndInfo(
"op-ctx", "my-ctx", "CONTEXT", null, null, Instant.now(), Instant.now(), null, null, false, null));
plugin.onInvocationEnd(new InvocationEndInfo("req-1", ARN, true, InvocationStatus.SUCCEEDED, null));

var operationSpan = spanByName(spanExporter.getFinishedSpanItems(), "my-ctx");
assertEquals(StatusCode.OK, operationSpan.getStatus().getStatusCode());
}

@Test
void operationNotCompleted_notEndedAtInvocationEnd() {
plugin.onInvocationStart(new InvocationInfo("req-1", ARN, true, Instant.now()));
Expand Down
96 changes: 96 additions & 0 deletions otel-plugin/src/test/java/software/amazon/lambda/durable/otel/InvocationOtelPluginTest.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -456,6 +456,102 @@ void userFunctionEnd_withFailure_setsErrorOnAttemptSpan() {
assertEquals(StatusCode.ERROR, attemptSpan.getStatus().getStatusCode());
}

@Test
void userFunctionEnd_withSuccess_setsOkOnAttemptSpan() {
plugin.onInvocationStart(new InvocationInfo("req-1", "arn:exec1", true, Instant.now()));

plugin.onUserFunctionStart(
new UserFunctionStartInfo("op-1", "compute", "STEP", "Step", null, Instant.now(), false, 1));
plugin.onUserFunctionEnd(new UserFunctionEndInfo(
"op-1", "compute", "STEP", "Step", null, Instant.now(), Instant.now(), false, 1, true, null));

plugin.onInvocationEnd(new InvocationEndInfo("req-1", "arn:exec1", true, InvocationStatus.SUCCEEDED, null));

var attemptSpan = spanExporter.getFinishedSpanItems().stream()
.filter(s -> s.getName().contains("attempt"))
.findFirst()
.orElseThrow();
assertEquals(StatusCode.OK, attemptSpan.getStatus().getStatusCode());
}

@Test
void operationEnd_withSuccess_setsOkOnOperationSpan() {
plugin.onInvocationStart(new InvocationInfo("req-1", "arn:exec1", true, Instant.now()));

plugin.onOperationStart(new OperationInfo("op-1", "step-ok", "STEP", "Step", null, Instant.now(), null, false));
plugin.onOperationEnd(new OperationEndInfo(
"op-1", "step-ok", "STEP", "Step", null, Instant.now(), Instant.now(), "SUCCEEDED", null, false, null));

plugin.onInvocationEnd(new InvocationEndInfo("req-1", "arn:exec1", true, InvocationStatus.SUCCEEDED, null));

var operationSpan = spanExporter.getFinishedSpanItems().stream()
.filter(s -> "step-ok".equals(s.getName()))
.findFirst()
.orElseThrow();
assertEquals(StatusCode.OK, operationSpan.getStatus().getStatusCode());
}

@Test
void operationEnd_withNonSuccessStatusAndNoError_leavesOperationSpanUnset() {
// onOperationEnd fires for every terminal status. A CANCELLED operation (or an error-less
// FAILED/TIMED_OUT/STOPPED) carries a non-null, non-SUCCEEDED status with a null error. It must NOT be
// stamped OK — the span status stays UNSET.
plugin.onInvocationStart(new InvocationInfo("req-1", "arn:exec1", true, Instant.now()));

plugin.onOperationStart(
new OperationInfo("op-cancel", "step-cancel", "STEP", "Step", null, Instant.now(), null, false));
plugin.onOperationEnd(new OperationEndInfo(
"op-cancel", "step-cancel", "STEP", "Step", null, Instant.now(), Instant.now(), "CANCELLED", null,
false, null));

plugin.onInvocationEnd(new InvocationEndInfo("req-1", "arn:exec1", true, InvocationStatus.SUCCEEDED, null));

var operationSpan = spanExporter.getFinishedSpanItems().stream()
.filter(s -> "step-cancel".equals(s.getName()))
.findFirst()
.orElseThrow();
assertEquals(StatusCode.UNSET, operationSpan.getStatus().getStatusCode());
}

@Test
void operationEnd_withoutMatchingStart_nonSuccessStatusAndNoError_leavesContinuationSpanUnset() {
// Same guard on the continuation-span branch (operation completed between invocations): an error-less
// TIMED_OUT terminal status must NOT be stamped OK.
plugin.onInvocationStart(new InvocationInfo("req-1", "arn:exec1", true, Instant.now()));

plugin.onOperationEnd(new OperationEndInfo(
"op-cb-timeout", "my-callback", "CALLBACK", "Callback", null, Instant.now(), Instant.now(),
"TIMED_OUT", null, false, null));

plugin.onInvocationEnd(new InvocationEndInfo("req-1", "arn:exec1", true, InvocationStatus.SUCCEEDED, null));

var continuationSpan = spanExporter.getFinishedSpanItems().stream()
.filter(s -> s.getName().contains("callback"))
.findFirst()
.orElseThrow();
assertEquals(StatusCode.UNSET, continuationSpan.getStatus().getStatusCode());
}

@Test
void operationEnd_withNullStatusAndNoError_setsOkOnOperationSpan() {
// A successful statusless virtual (FLAT CONTEXT) operation fires onOperationEnd with a null operation ->
// null status and null error. This is genuine success and must be stamped OK.
plugin.onInvocationStart(new InvocationInfo("req-1", "arn:exec1", true, Instant.now()));

plugin.onOperationStart(
new OperationInfo("op-ctx", "my-ctx", "CONTEXT", null, null, Instant.now(), null, false));
plugin.onOperationEnd(new OperationEndInfo(
"op-ctx", "my-ctx", "CONTEXT", null, null, Instant.now(), Instant.now(), null, null, false, null));

plugin.onInvocationEnd(new InvocationEndInfo("req-1", "arn:exec1", true, InvocationStatus.SUCCEEDED, null));

var operationSpan = spanExporter.getFinishedSpanItems().stream()
.filter(s -> "my-ctx".equals(s.getName()))
.findFirst()
.orElseThrow();
assertEquals(StatusCode.OK, operationSpan.getStatus().getStatusCode());
}

@Test
void fullLifecycle_producesCorrectSpanHierarchy() {
var arn = "arn:aws:lambda:us-east-1:123:function:test:$LATEST/durable/exec1";
Expand Down
Loading