yujunjun/datahub
2
0
Fork 0
mirror of https://github.com/datahub-project/datahub.git synced 2025-07-03 23:28:11 +00:00
Code Issues Packages Projects Releases Wiki Activity
datahub/docs/api/tutorials/assertions.md

1193 lines
28 KiB
Markdown
Raw Blame History

import Tabs from '@theme/Tabs';
import TabItem from '@theme/TabItem';
# Assertions
<FeatureAvailability saasOnly />
This guide specifically covers how to use the Assertion APIs for **DataHub Cloud** native assertions, including:
- [Freshness Assertions](/docs/managed-datahub/observe/freshness-assertions.md)
- [Volume Assertions](/docs/managed-datahub/observe/volume-assertions.md)
- [Column Assertions](/docs/managed-datahub/observe/column-assertions.md)
- [Schema Assertions](/docs/managed-datahub/observe/schema-assertions.md)
- [Custom SQL Assertions](/docs/managed-datahub/observe/custom-sql-assertions.md)
## Why Would You Use Assertions APIs?
The Assertions APIs allow you to create, schedule, run, and delete Assertions with DataHub Cloud. Additionally, you can manage subscriptions to receive notifications when assertions change state or when other entity changes occur.
### Goal Of This Guide
This guide will show you how to create, schedule, run and delete Assertions for a Table.
## Prerequisites
The actor making API calls must have the `Edit Assertions` and `Edit Monitors` privileges for the Tables at hand.
## Create Assertions
You can create new dataset Assertions to DataHub using the following APIs.
<Tabs>
<TabItem value="graphql" label="GraphQL" default>
### Freshness Assertion
To create a new freshness assertion, use the `upsertDatasetFreshnessAssertionMonitor` GraphQL Mutation.
```graphql
mutation upsertDatasetFreshnessAssertionMonitor {
upsertDatasetFreshnessAssertionMonitor(
input: {
entityUrn: "<urn of entity being monitored>"
schedule: {
type: FIXED_INTERVAL
fixedInterval: { unit: HOUR, multiple: 8 }
}
evaluationSchedule: {
timezone: "America/Los_Angeles"
cron: "0 */8 * * *"
}
evaluationParameters: { sourceType: INFORMATION_SCHEMA }
mode: ACTIVE
}
) {
urn
}
}
```
This API will return a unique identifier (URN) for the new assertion if you were successful:
```json
{
"data": {
"upsertDatasetFreshnessAssertionMonitor": {
"urn": "urn:li:assertion:your-new-assertion-id"
}
},
"extensions": {}
}
```
For more details, see the [Freshness Assertions](/docs/managed-datahub/observe/freshness-assertions.md) guide.
### Volume Assertions
To create a new volume assertion, use the `upsertDatasetVolumeAssertionMonitor` GraphQL Mutation.
```graphql
mutation upsertDatasetVolumeAssertionMonitor {
upsertDatasetVolumeAssertionMonitor(
input: {
entityUrn: "<urn of entity being monitored>"
type: ROW_COUNT_TOTAL
rowCountTotal: {
operator: BETWEEN
parameters: {
minValue: { value: "10", type: NUMBER }
maxValue: { value: "20", type: NUMBER }
}
}
evaluationSchedule: {
timezone: "America/Los_Angeles"
cron: "0 */8 * * *"
}
evaluationParameters: { sourceType: INFORMATION_SCHEMA }
mode: ACTIVE
}
) {
urn
}
}
```
This API will return a unique identifier (URN) for the new assertion if you were successful:
```json
{
"data": {
"upsertDatasetVolumeAssertionMonitor": {
"urn": "urn:li:assertion:your-new-assertion-id"
}
},
"extensions": {}
}
```
For more details, see the [Volume Assertions](/docs/managed-datahub/observe/volume-assertions.md) guide.
### Column Assertions
To create a new column assertion, use the `upsertDatasetFieldAssertionMonitor` GraphQL Mutation.
```graphql
mutation upsertDatasetFieldAssertionMonitor {
upsertDatasetFieldAssertionMonitor(
input: {
entityUrn: "<urn of entity being monitored>"
type: FIELD_VALUES
fieldValuesAssertion: {
field: {
path: "<name of the column to be monitored>"
type: "NUMBER"
nativeType: "NUMBER(38,0)"
}
operator: GREATER_THAN
parameters: { value: { type: NUMBER, value: "10" } }
failThreshold: { type: COUNT, value: 0 }
excludeNulls: true
}
evaluationSchedule: {
timezone: "America/Los_Angeles"
cron: "0 */8 * * *"
}
evaluationParameters: { sourceType: ALL_ROWS_QUERY }
mode: ACTIVE
}
) {
urn
}
}
```
This API will return a unique identifier (URN) for the new assertion if you were successful:
```json
{
"data": {
"upsertDatasetFieldAssertionMonitor": {
"urn": "urn:li:assertion:your-new-assertion-id"
}
},
"extensions": {}
}
```
For more details, see the [Column Assertions](/docs/managed-datahub/observe/column-assertions.md) guide.
### Custom SQL Assertions
To create a new column assertion, use the `upsertDatasetSqlAssertionMonitor` GraphQL Mutation.
```graphql
mutation upsertDatasetSqlAssertionMonitor {
upsertDatasetSqlAssertionMonitor(
assertionUrn: "<urn of assertion created in earlier query>"
input: {
entityUrn: "<urn of entity being monitored>"
type: METRIC
description: "<description of the custom assertion>"
statement: "<SQL query to be evaluated>"
operator: GREATER_THAN_OR_EQUAL_TO
parameters: { value: { value: "100", type: NUMBER } }
evaluationSchedule: {
timezone: "America/Los_Angeles"
cron: "0 */6 * * *"
}
mode: ACTIVE
}
) {
urn
}
}
```
This API will return a unique identifier (URN) for the new assertion if you were successful:
```json
{
"data": {
"upsertDatasetSqlAssertionMonitor": {
"urn": "urn:li:assertion:your-new-assertion-id"
}
},
"extensions": {}
}
```
For more details, see the [Custom SQL Assertions](/docs/managed-datahub/observe/custom-sql-assertions.md) guide.
### Schema Assertions
To create a new schema assertion, use the `upsertDatasetSchemaAssertionMonitor` GraphQL Mutation.
```graphql
mutation upsertDatasetSchemaAssertionMonitor {
upsertDatasetSchemaAssertionMonitor(
assertionUrn: "urn:li:assertion:existing-assertion-id"
input: {
entityUrn: "<urn of the table to be monitored>"
assertion: {
compatibility: EXACT_MATCH
fields: [
{ path: "id", type: STRING }
{ path: "count", type: NUMBER }
{ path: "struct", type: STRUCT }
{ path: "struct.nestedBooleanField", type: BOOLEAN }
]
}
description: "<description of the schema assertion>"
mode: ACTIVE
}
)
}
```
This API will return a unique identifier (URN) for the new assertion if you were successful:
```json
{
"data": {
"upsertDatasetSchemaAssertionMonitor": {
"urn": "urn:li:assertion:your-new-assertion-id"
}
},
"extensions": {}
}
```
For more details, see the [Schema Assertions](/docs/managed-datahub/observe/schema-assertions.md) guide.
</TabItem>
</Tabs>
## Run Assertions
You can use the following APIs to trigger the assertions you've created to run on-demand. This is
particularly useful for running assertions on a custom schedule, for example from your production
data pipelines.
> **Long-Running Assertions**: The timeout for synchronously running an assertion is currently limited to a maximum of 30 seconds.
> Each of the following APIs support an `async` parameter, which can be set to `true` to run the assertion asynchronously.
> When set to `true`, the API will kick off the assertion run and return null immediately. To view the result of the assertion,
> simply fetching the runEvents field of the `assertion(urn: String!)` GraphQL query.
<Tabs>
<TabItem value="graphql" label="GraphQL" default>
### Run Assertion
```graphql
mutation runAssertion {
runAssertion(urn: "urn:li:assertion:your-assertion-id", saveResult: true) {
type
nativeResults {
key
value
}
}
}
```
Where **type** will contain the Result of the assertion run, either `SUCCESS`, `FAILURE`, or `ERROR`.
The `saveResult` argument determines whether the result of the assertion will be saved to DataHub's backend,
and available to view through the DataHub UI. If this is set to false, the result will NOT be stored in DataHub's
backend. The value defaults to `true`.
If the assertion is external (not natively executed by DataHub), this API will return an error.
If running the assertion is successful, the result will be returned as follows:
```json
{
"data": {
"runAssertion": {
"type": "SUCCESS",
"nativeResults": [
{
"key": "Value",
"value": "1382"
}
]
}
},
"extensions": {}
}
```
### Run Group of Assertions
```graphql
mutation runAssertions {
runAssertions(
urns: [
"urn:li:assertion:your-assertion-id-1"
"urn:li:assertion:your-assertion-id-2"
]
saveResults: true
) {
passingCount
failingCount
errorCount
results {
urn
result {
type
nativeResults {
key
value
}
}
}
}
}
```
Where **type** will contain the Result of the assertion run, either `SUCCESS`, `FAILURE`, or `ERROR`.
The `saveResults` argument determines whether the result of the assertion will be saved to DataHub's backend,
and available to view through the DataHub UI. If this is set to false, the result will NOT be stored in DataHub's
backend. The value defaults to `true`.
If any of the assertion are external (not natively executed by DataHub), they will simply be omitted from the result set.
If running the assertions is successful, the results will be returned as follows:
```json
{
"data": {
"runAssertions": {
"passingCount": 2,
"failingCount": 0,
"errorCount": 0,
"results": [
{
"urn": "urn:li:assertion:your-assertion-id-1",
"result": {
"type": "SUCCESS",
"nativeResults": [
{
"key": "Value",
"value": "1382"
}
]
}
},
{
"urn": "urn:li:assertion:your-assertion-id-2",
"result": {
"type": "FAILURE",
"nativeResults": [
{
"key": "Value",
"value": "12323"
}
]
}
}
]
}
},
"extensions": {}
}
```
Where you should see one result object for each assertion.
### Run All Assertions for Table
You can also run all assertions for a specific data asset using the `runAssertionsForAsset` mutation.
```graphql
mutation runAssertionsForAsset {
runAssertionsForAsset(
urn: "urn:li:dataset:(urn:li:dataPlatform:snowflake,purchase_events,PROD)"
saveResults: true
) {
passingCount
failingCount
errorCount
results {
urn
result {
type
nativeResults {
key
value
}
}
}
}
}
```
Where `type` will contain the Result of the assertion run, either `SUCCESS`, `FAILURE`, or `ERROR`.
The `saveResults` argument determines whether the result of the assertion will be saved to DataHub's backend,
and available to view through the DataHub UI. If this is set to false, the result will NOT be stored in DataHub's
backend. The value defaults to `true`.
If any of the assertion are external (not natively executed by DataHub), they will simply be omitted from the result
set.
If running the assertions is successful, the results will be returned as follows:
```json
{
"data": {
"runAssertionsForAsset": {
"passingCount": 2,
"failingCount": 0,
"errorCount": 0,
"results": [
{
"urn": "urn:li:assertion:your-assertion-id-1",
"result": {
"type": "SUCCESS",
"nativeResults": [
{
"key": "Value",
"value": "1382"
}
]
}
},
{
"urn": "urn:li:assertion:your-assertion-id-2",
"result": {
"type": "FAILURE",
"nativeResults": [
{
"key": "Value",
"value": "12323"
}
]
}
}
]
}
},
"extensions": {}
}
```
Where you should see one result object for each assertion.
### Run Group of Assertions for Table
If you don't always want to run _all_ assertions for a given table, you can also opt to run a subset of the
table's assertions using _Assertion Tags_. First, you'll add tags to your assertions to group and categorize them,
then you'll call the `runAssertionsForAsset` mutation with the `tagUrns` argument to filter for assertions having those tags.
#### Step 1: Adding Tag to an Assertion
Currently, you can add tags to an assertion only via the DataHub GraphQL API. You can do this using the following mutation:
```graphql
mutation addTags {
addTag(
input: {
resourceUrn: "urn:li:assertion:your-assertion"
tagUrn: "urn:li:tag:my-important-tag"
}
)
}
```
#### Step 2: Run All Assertions for a Table with Tags
Now, you can run all assertions for a table with a specific tag(s) using the `runAssertionsForAsset` mutation with the
`tagUrns` input parameter:
```graphql
mutation runAssertionsForAsset {
runAssertionsForAsset(
urn: "urn:li:dataset:(urn:li:dataPlatform:snowflake,purchase_events,PROD)"
tagUrns: ["urn:li:tag:my-important-tag"]
) {
passingCount
failingCount
errorCount
results {
urn
result {
type
nativeResults {
key
value
}
}
}
}
}
```
**Coming Soon**: Support for adding tags to assertions through the DataHub UI.
</TabItem>
<TabItem value="python" label="Python">
### Run Assertion
```python
{{ inline /metadata-ingestion/examples/library/run_assertion.py show_path_as_comment }}
```
### Run Group of Assertions
```python
{{ inline /metadata-ingestion/examples/library/run_assertions.py show_path_as_comment }}
```
### Run All Assertions for Table
```python
{{ inline /metadata-ingestion/examples/library/run_assertions_for_asset.py show_path_as_comment }}
```
</TabItem>
</Tabs>
### Experimental: Providing Dynamic Parameters to Assertions
You can provide **dynamic parameters** to your assertions to customize their behavior. This is particularly useful for
assertions that require dynamic parameters, such as a threshold value that changes based on the time of day.
Dynamic parameters can be injected into the SQL fragment portion of any Assertion. For example, it can appear
in any part of the SQL statement in a [Custom SQL](/docs/managed-datahub/observe/custom-sql-assertions.md) Assertion,
or it can appear in the **Advanced > Filter** section of a [Column](/docs/managed-datahub/observe/column-assertions.md),
[Volume](/docs/managed-datahub/observe/volume-assertions.md), or [Freshness](/docs/managed-datahub/observe/freshness-assertions.md) Assertion.
To do so, you'll first need to edit the SQL fragment to include the dynamic parameter. Dynamic parameters appear
as `${parameterName}` in the SQL fragment.
Next, you'll call the `runAssertion`, `runAssertions`, or `runAssertionsForAsset` mutations with the `parameters` input argument.
This argument is a list of key-value tuples, where the key is the parameter name and the value is the parameter value:
```graphql
mutation runAssertion {
runAssertion(
urn: "urn:li:assertion:your-assertion-id"
parameters: [{ key: "parameterName", value: "parameterValue" }]
) {
type
nativeResults {
key
value
}
}
}
```
At runtime, the `${parameterName}` placeholder in the SQL fragment will be replaced with the provided `parameterValue` before the query
is sent to the database for execution.
## Get Assertion Details
You can use the following APIs to
1. Fetch existing assertion definitions + run history
2. Fetch the assertions associated with a given table + their run history.
<Tabs>
<TabItem value="graphql" label="GraphQL" default>
### Get Assertions for Table
To retrieve all the assertions for a table, you can use the following GraphQL Query.
```graphql
query dataset {
dataset(
urn: "urn:li:dataset:(urn:li:dataPlatform:snowflake,purchases,PROD)"
) {
assertions(start: 0, count: 1000) {
start
count
total
assertions {
urn
# Fetch the last run of each associated assertion.
runEvents(status: COMPLETE, limit: 1) {
total
failed
succeeded
runEvents {
timestampMillis
status
result {
type
nativeResults {
key
value
}
}
}
}
info {
type
description
lastUpdated {
time
actor
}
datasetAssertion {
datasetUrn
scope
aggregation
operator
parameters {
value {
value
type
}
minValue {
value
type
}
maxValue {
value
type
}
}
fields {
urn
path
}
nativeType
nativeParameters {
key
value
}
logic
}
freshnessAssertion {
type
entityUrn
schedule {
type
cron {
cron
timezone
}
fixedInterval {
unit
multiple
}
}
filter {
type
sql
}
}
sqlAssertion {
type
entityUrn
statement
changeType
operator
parameters {
value {
value
type
}
minValue {
value
type
}
maxValue {
value
type
}
}
}
fieldAssertion {
type
entityUrn
filter {
type
sql
}
fieldValuesAssertion {
field {
path
type
nativeType
}
transform {
type
}
operator
parameters {
value {
value
type
}
minValue {
value
type
}
maxValue {
value
type
}
}
failThreshold {
type
value
}
excludeNulls
}
fieldMetricAssertion {
field {
path
type
nativeType
}
metric
operator
parameters {
value {
value
type
}
minValue {
value
type
}
maxValue {
value
type
}
}
}
}
volumeAssertion {
type
entityUrn
filter {
type
sql
}
rowCountTotal {
operator
parameters {
value {
value
type
}
minValue {
value
type
}
maxValue {
value
type
}
}
}
rowCountChange {
type
operator
parameters {
value {
value
type
}
minValue {
value
type
}
maxValue {
value
type
}
}
}
}
schemaAssertion {
entityUrn
compatibility
fields {
path
type
nativeType
}
schema {
fields {
fieldPath
type
nativeDataType
}
}
}
source {
type
created {
time
actor
}
}
}
}
}
}
}
```
### Get Assertion Details
You can use the following GraphQL query to fetch the details for an assertion along with its evaluation history by URN.
```graphql
query getAssertion {
assertion(urn: "urn:li:assertion:assertion-id") {
urn
# Fetch the last 10 runs for the assertion.
runEvents(status: COMPLETE, limit: 10) {
total
failed
succeeded
runEvents {
timestampMillis
status
result {
type
nativeResults {
key
value
}
}
}
}
info {
type
description
lastUpdated {
time
actor
}
datasetAssertion {
datasetUrn
scope
aggregation
operator
parameters {
value {
value
type
}
minValue {
value
type
}
maxValue {
value
type
}
}
fields {
urn
path
}
nativeType
nativeParameters {
key
value
}
logic
}
freshnessAssertion {
type
entityUrn
schedule {
type
cron {
cron
timezone
}
fixedInterval {
unit
multiple
}
}
filter {
type
sql
}
}
sqlAssertion {
type
entityUrn
statement
changeType
operator
parameters {
value {
value
type
}
minValue {
value
type
}
maxValue {
value
type
}
}
}
fieldAssertion {
type
entityUrn
filter {
type
sql
}
fieldValuesAssertion {
field {
path
type
nativeType
}
transform {
type
}
operator
parameters {
value {
value
type
}
minValue {
value
type
}
maxValue {
value
type
}
}
failThreshold {
type
value
}
excludeNulls
}
fieldMetricAssertion {
field {
path
type
nativeType
}
metric
operator
parameters {
value {
value
type
}
minValue {
value
type
}
maxValue {
value
type
}
}
}
}
volumeAssertion {
type
entityUrn
filter {
type
sql
}
rowCountTotal {
operator
parameters {
value {
value
type
}
minValue {
value
type
}
maxValue {
value
type
}
}
}
rowCountChange {
type
operator
parameters {
value {
value
type
}
minValue {
value
type
}
maxValue {
value
type
}
}
}
}
schemaAssertion {
entityUrn
compatibility
fields {
path
type
nativeType
}
schema {
fields {
fieldPath
type
nativeDataType
}
}
}
source {
type
created {
time
actor
}
}
}
}
}
```
</TabItem>
<TabItem value="python" label="Python">
```python
Python support coming soon!
```
</TabItem>
</Tabs>
## Add Tag to Assertion
You can add tags to individual assertions to group and categorize them, for example by its priority or severity.
Note that the tag should already exist in DataHub, or the operation will fail.
<Tabs>
<TabItem value="graphql" label="GraphQL" default>
```graphql
mutation addTags {
addTag(
input: {
resourceUrn: "urn:li:assertion:your-assertion"
tagUrn: "urn:li:tag:my-important-tag"
}
)
}
```
If you see the following response, the operation was successful:
```json
{
"data": {
"addTag": true
},
"extensions": {}
}
```
You can create new tags using the `createTag` mutation or via the UI.
</TabItem>
</Tabs>
## Delete Assertions
You can use delete dataset operations to DataHub using the following APIs.
<Tabs>
<TabItem value="graphql" label="GraphQL" default>
```graphql
mutation deleteAssertion {
deleteAssertion(urn: "urn:li:assertion:test")
}
```
If you see the following response, the operation was successful:
```json
{
"data": {
"deleteAssertion": true
},
"extensions": {}
}
```
</TabItem>
<TabItem value="python" label="Python">
```python
{{ inline /metadata-ingestion/examples/library/delete_assertion.py show_path_as_comment }}
```
</TabItem>
</Tabs>
## (Advanced) Create and Report Results for Custom Assertions
If you'd like to create and report results for your own custom assertions, e.g. those which are run and
evaluated outside of DataHub Cloud, you need to generate 2 important Assertion Entity aspects, and give the assertion a unique
URN of the following format:
1. Generate a unique URN for your assertion
```plaintext
urn:li:assertion:<unique-assertion-id>
```
2. Generate the [**AssertionInfo**](/docs/generated/metamodel/entities/assertion.md#assertion-info) aspect for the assertion. You can do this using the Python SDK. Give your assertion a `type` and a `source`
with type `EXTERNAL` to mark it as an external assertion, not run by DataHub itself.
3. Generate the [**AssertionRunEvent**](/docs/generated/metamodel/entities/assertion.md#assertionrunevent-timeseries) timeseries aspect using the Python SDK. This aspect should contain the result of the assertion
run at a given timestamp and will be shown on the results graph in DataHub's UI.
## Create Subscription
You can create subscriptions to receive notifications when assertions change state (pass, fail, or error) or when other entity changes occur. Subscriptions can be created at the dataset level (affecting all assertions on the dataset) or at the assertion level (affecting only specific assertions).
<Tabs>
<TabItem value="python" label="Python">
```python
{{ inline /metadata-ingestion/examples/library/create_subscription.py show_path_as_comment }}
```
</TabItem>
</Tabs>
## Remove Subscription
You can remove existing subscriptions to stop receiving notifications. The unsubscribe method supports selective removal of specific change types or complete removal of subscriptions.
<Tabs>
<TabItem value="python" label="Python">
```python
{{ inline /metadata-ingestion/examples/library/remove_subscription.py show_path_as_comment }}
```
</TabItem>
</Tabs>
Reference in New Issue View Git Blame Copy Permalink