Loosen LtsMigrationService.select() to pick every bean in the (from, to]
range regardless of family, so a 4.3.x -> 4.4 offline upgrade runs the real
in-range 4.3.1.x beans plus a new 4.4.0.0 baseline bean, each exactly once.
This replaces the ad-hoc explicit V4_3_1_3Migration.apply() call in
DefaultDataUpdateService and moves the 4.4 baseline flat DDL out of
basic/schema_update.sql into V4_4_0_0Migration / lts/4.4.0.0/schema_update.sql.
- Drop the dangling "spec §7" reference from the test comments; inline the
actual rule (a lexical MIN/MAX result over string telemetry stays a string).
- Use a zero-padded code ("0009") for the string-result case - a clearer example
of a genuine string that must not be coerced to a number (would lose padding).
- Define the maxCode metric in the test CF configuration alongside consumption
and avgConsumption, so all parameterized metric names are configured.
Address review feedback on the integration test:
- Read latest telemetry with useStrictDataTypes=true always, so the value node
keeps its stored type (numeric -> JSON number, str_v -> JSON string).
- Assert the aggregation result is a numeric node across all existing scenarios
via a shared assertNumericValue helper, instead of a separate standalone test.
isNumber() is the actual regression guard; asLong()/asText() would coerce a
str_v string and miss it, so the type is asserted explicitly and the value is
compared numerically.
- Drop the redundant standalone test that duplicated an existing scenario.
Cover the fix that makes EntityAggregationCalculatedFieldState.toResult()
emit numeric aggregation results as numeric JSON nodes (dbl_v/long_v) instead
of JSON strings (str_v), which had broken server-side AVG/SUM aggregation.
- New unit test EntityAggregationCalculatedFieldStateTest: asserts toResult()
serializes a numeric result as a numeric node and a genuine string result
(lexical MIN/MAX over string telemetry) as a string node. Assertions check
the node type (isNumber/isTextual) rather than asText(), since asText()
coerces both node kinds identically - which is why the bug went unnoticed.
- EntityAggregationCalculatedFieldTest: add a strict-types read helper
(useStrictDataTypes=true) and a test asserting the stored SUM/AVG telemetry
are numeric JSON nodes end-to-end. Existing asText()-based assertions are
left untouched (they read via the non-strict endpoint, which stringifies
every value and therefore cannot observe the storage type).
EntityAggregationCalculatedFieldState.toResult() serialized every metric
result via ObjectNode.put(name, JacksonUtil.toString(value)), which produces
a JSON *string* node even for numeric aggregation results (SUM/AVG/COUNT/...).
When persisted (in particular with transport.json.type_cast_enabled=false, or
for non-parsable values) the result lands in ts_kv.str_v, so server-side AVG/SUM
aggregation returns no data in widgets/queries while COUNT/MIN/MAX still return.
Serialize with JacksonUtil.valueToTree(...) so numeric results are emitted as
numeric JSON nodes (-> dbl_v/long_v), mirroring
RelatedEntitiesAggregationCalculatedFieldState. Genuine string MIN/MAX results
over string telemetry are preserved as string nodes. The instanceof Number
guard is retained (it only governs rounding).
A device-state attribute save can complete after its tenant was deleted.
The post-save WS-update callback then resolves an evicted tenant profile
via partitionService.resolve -> getRoutingInfo and throws
TenantNotFoundException uncaught on the ws-callback thread.
Guard resolve() in forwardToSubscriptionManagerService: a deleted tenant
has no partition to route to and no subscribers to notify, so skip the
forward instead of propagating. Mirrors the existing tenant-gone handling
in DefaultDeviceStateService.checkStates().
The deprecation cleanup set @EqualsAndHashCode(callSuper = true) on AlarmComment,
so equals/hashCode now include the BaseData id and createdTime. The delete tests
built the expected comment via a fresh builder (no id), which no longer matches the
server-generated deletion comment (which reuses the original comment id).
Mutate the original alarm comment (already carrying the correct id/createdTime) into
the expected SYSTEM deletion comment instead of building a fresh one.
* feat: send pseudonymized install report (tenant/user hash, tb version) to IoT Hub
* refactor: source install-report tb version/edition from ProjectInfo
Use the existing ProjectInfo component (getProjectVersion/getProductType)
instead of an ad-hoc optional BuildProperties field and a hard-coded "CE"
edition, so PE reports its own product type and the version is the canonical
cleaned value. Constructor-injected via @RequiredArgsConstructor.
If iotHubInstalledItemService.save throws after the local entity was already created, the entity was orphaned — visible in the tenant but invisible to IoT Hub update/delete. Extracted the per-type entity-deletion dispatch from deleteInstalledItem into a deleteEntityForDescriptor helper and call it from doInstallVersion and registerDeviceInstall on tracking-save failure. registerDeviceInstall now also wraps reportVersionInstalled best-effort to match doInstallVersion.
The post-install Thread.sleep was driven by installTimeoutMs from the uploaded solution.json with no upper bound — a malicious or misconfigured template could pin an HTTP worker thread indefinitely. Cap it at a configurable max (default 60s) exposed as iot-hub.max-install-timeout-ms with an IOT_HUB_MAX_INSTALL_TIMEOUT_MS env override.
Stream entries through a fixed buffer and abort with IOException when the archive crosses any of three configurable thresholds: total uncompressed bytes, per-entry uncompressed bytes, or entry count. Defaults — 200 MiB total, 50 MiB per entry, 10 000 entries — and env-var overrides IOT_HUB_MAX_UNCOMPRESSED_ARCHIVE_BYTES / IOT_HUB_MAX_UNCOMPRESSED_ENTRY_BYTES / IOT_HUB_MAX_ARCHIVE_ENTRY_COUNT are exposed under iot-hub in thingsboard.yml.
Root entries other than WIDGET or SOLUTION_TEMPLATE should always be installed even when a prior version is already installed — only widgets and solution templates are deduplicated by item id at install-plan resolution time.
Route all parse-failure throws through a single parseFailure(action, itemType[, section], cause) helper so the user-facing wording lives in one place and update paths share the template with install paths.
Andrii Landiak
Viacheslav Klimov
Nikita Mazurenko
dshvaika
Oleksandra Matviienko
IrynaMatveieva
Andrew Shvayka
Igor Kulikov
Maksym Tsymbarov
dpinkevych