Business Intelligence API
The Business Intelligence API allows to get audit logs from the rule solver.
The Business Intelligence API is providing the business intelligence functionalities of DecisionRules. Most importantly, it allows to collect data from the rule solver.
The API allows you to get audits from database and delete them from the database when they are no longer needed. We describe both endpoints below.
This endpoint allows you to fetch audit logs from your rule solver, including metadata of the solver run as well as the input and output data.
The rule solver does not generate audit logs by default. If you want some rule to generate audits on each solve, you have to turn on audit logs in its Rule Settings. More information can be found on the Audit Logging Page.
get
https://bi.decisionrules.io/
audit
Get audits the from rule solver
URL
https://bi.decisionrules.io/audit?limit=1
Headers:
Content-Type: application/json
Authorization: Bearer DOZpz-h6xnOrKGIINlYvkd9hn41pRR3oG6cqH
You need to copy paste your own Business Intelligence API Key after
Bearer . If you do not have it yet, generate your Business Intelligence API Key in the app.If you cannot send the Business Intelligence API Key inside the authorization header, you may use the
bi_key query parameter; e.g. bi_key=DOZpz-h6xnOrKGIINlYvkd9hn41pRR3oG6cqH. In that case, the request could look as follows.URL
https://bi.decisionrules.io/audit?limit=1&bi_key=DOZpz-h6xnOrKGIINlYvkd9hn41pRR3oG6cqH
Headers:
Content-Type: application/json
If the
bi_key query parameter is not set, the server will look for the authorization header as described above.{
"audits": [
{
"correlationId": "7bac2e34-53c6-cbb0-d898-87f46ad50f77",
"guid": "b88ba671-eae9-05e8-a547-f54d0cdedba7",
"baseId": "d58e2518-57f6-a315-2458-b931b45cb543",
"version": 1,
"type": "decision-table",
"status": "published",
"outputSchema": {
"finalPrice": {},
"message": {}
},
"createdIn": "2022-06-28T13:22:11.875Z",
"inputSchema": {
"benefitCode": {},
"promoCode": {},
"price": {}
},
"tags": [
"Benefit"
],
"name": "Benefit Program",
"lastUpdate": "2022-06-28T13:22:11.875Z",
"solverKey": "QyjvIdGhdDgUP6_Lebo2DOyYaoSIA0axZlVcRrxbJzHazTOMn03GA5kiuIzgIw5z",
"inputData": {
"benefitCode": "SUMMER",
"promoCode": {},
"price": {}
},
"outputData": [
{
"finalPrice": {},
"message": "discount 30%"
},
{
"finalPrice": {},
"message": "no discount"
}
],
"executionTime": 1,
"statusCode": 200,
"errorMessage": null,
"debugData": null,
"timestamp": "2022-06-28T13:22:28.579Z"
}
]
}
There can be a lot of audit logs in the database, of course. We therefore enforce a maximum page size, e.g. the number of audit logs that can be returned per one request. The maximum page size is 1.000, which means that you can never get more audits than 1.000 per one call.
You can choose a custom page size smaller than 1.000 by defining the
page_size query parameter. For example, if you set page_size=100, the API will always return maximum 100 audits per request, no matter how many matching audits there are in the database. If not set, the page size defaults to 1.000.To get more audit logs than those that can fit on one page, you can use the
page parameter. For example, if you set page_size=100&page=2, you will get audits 101 - 200, while a request with page_size=100&page=3 yields audits 201 - 300. etc. If not set, page defaults to 1 so you get the first page.You can also specify the total limit on the number of audits by setting the
limit parameter. If set, the API will never fetch any further audits than that with index equal to the limit. For example, a request with page_size=100&page=2&limit=150 yields only the audits 101 - 150, while page_size=100&page=3&limit=150 returns just an empty array of audits because you have exceeded your limit. There is no limit by default.If you are collecting debug data, you may choose to include them in your audit logs. This can be done with the
include_debug query parameter. If set to true via include_debug=true, this query parameter tells the server to include debug data in your audit logs, whenever applicable.The rest of the query parameters work as filters that help you specify the desired audits.
When calling the Rule Solver API, the solver provides the audit log with the so-called correlation ID. This can happen in two distinct ways. If the request to the solver contains the X-Correlation-Id header, the value of this header is simply passed to the correlation ID of the audit log. If the request to the solver does not contain such header, the solver generates its own (unique) correlation ID.
Note that, in both cases, the solver returns the correlation ID in the X-Correlation-Id header of the response.
The correlation ID can then be used to get the desired audit log related to a particular solver request. This can be done easily by including the
correlation_ids parameter. It takes an array of correlation IDs separated by a comma, e.g. ., correlation_ids=af4012bd-d492-92ec-ffa4-31fd2b70b1bc,197d5d5a-f6f7-35de-1afb-dc26237ebfc9.If set, the
rules parameter limits the audits to those produced by the specified rules. The individual rules are identified by ids separated by comma, i.e., rules=af4012bd-d492-92ec-ffa4-31fd2b70b1bc,197d5d5a-f6f7-35de-1afb-dc26237ebfc9. Moreover, one may further specify the allowed versions for each of the rules by including square brackets with the comma separated list of versions. For instance, rules=af4012bd-d492-92ec-ffa4-31fd2b70b1bc[1,2],197d5d5a-f6f7-35de-1afb-dc26237ebfc9 will return audits for rules with id af4012bd-d492-92ec-ffa4-31fd2b70b1bc whose version is either 1 or 2 and audits for rules with id 197d5d5a-f6f7-35de-1afb-dc26237ebfc9 of any version.It is also possible to filter solemnly by rule versions. To do that, use the square bracket expression without specifying rule IDs. For instance,
rules=[1,2] will return audit logs for rules whose version is 1 or 2, no matter the rule ID.If set, the
solver_keys parameter limits the audits to those produced by calling rules with the specified Solver API Key. The individual keys are separated by comma.The
tags query parameter allows you to filter the audits by tags present on the rules. For the audit to pass the filter, its rule needs to be decorated with all the tags specified in the query parameter. The tags are separated by comma. For example, a request with tags=Pricing,Test returns only audits from rules decorated by both the Pricing and the Test tag.The
date_gte and date_lte parameters allow to filter audit logs by date (and time). The former serves as a lower bound on the date of the solve, the latter serves as an upper bound. In other words, audits will match this filter only if they are older than date_gte and younger than date_lte. When specifying one or both of the parameters, you can choose from a number of supported formats, including ISO 8601 and RFC 2822 Date time. For example:2013-02-08
2013-02
2013-02-08 09
2013-02-08 09:30
2013-02-08 09:30:26.123
2013-02-08 09+07:00
2013-02-08 09Z
2013-02-08 09:30:26.123+07:00
...
6 Mar 17 21:22 UT
d6 Mar 2017 21:22:23 GMT
Mon, 06 Mar 2017 21:22:23 +0000
...
Note that special characters have to be URL encoded when passed to the query parameter. Therefore, when writing your date, please replace these characters with their respective codes. We list the most useful characters and their codes below.
space %20
: %3A
+ %2B
The
order parameter can take two values, order=asc or order=desc. It specifies the chronological order of the audits returned. For order=asc, the audits are sorted chronologically, starting with the oldest. For order=desc, they are sorted chronologically, starting with the latest. By default, the audits are returned in ascending order.If set, the
status_codes parameter limits the audit logs to those produced with the specified status code. The individual codes are separated by comma. For example, to get logs from rule solves that returned OK, use status_codes=200. To get audit logs from rule solves that returned some kind of rule error, use status_codes=400, 401, 404, 406, 426. Eventually, to get audit logs from rule solves that returned server error, use status_codes=500.Finally let us give an example of a request to the Business Intelligence API with a few optional query parameters.
URL
https://bi.decisionrules.io/audit?page_size=100&page=2&tags=Pricing&date_gte=2022-01-01&date_lte=2022-31-12
Headers:
Content-Type: application/json
Authorization: Bearer DOZpz-h6xnOrKGIINlYvkd9hn41pRR3oG6cqH
You need to copy paste your own Business Intelligence API Key after
Bearer . If you do not have it yet, generate your Business Intelligence API Key in the app.The above request will get audit logs from all solver runs in the given time period (year 2022) of all rules with tag Pricing . We specify that the page size is 100 and we want the second page, i.e., we get audits 101 - 200 (if they are in the database, of course).
This endpoint allows you to delete audit logs from the database when you no longer need them.
This action results in permanent and irreversible disposal of the deleted audit logs. Please, use it carefully. More information can be found on the Audit Logging page.
delete
https://bi.decisionrules.io/
audit
Delete audits from the database
URL
https://bi.decisionrules.io/audit?limit=1
Headers:
Content-Type: application/json
Authorization: Bearer DOZpz-h6xnOrKGIINlYvkd9hn41pRR3oG6cqH
You need to copy paste your own Business Intelligence API Key after
Bearer . If you do not have it yet, generate your Business Intelligence API Key in the app.The Business Intelligence API Key can be also included in a query parameter instead of the header. In that case, the same request would look as follows.
URL
https://bi.decisionrules.io/audit?limit=1&bi_key=DOZpz-h6xnOrKGIINlYvkd9hn41pRR3oG6cqH
Headers:
Content-Type: application/json
Just as in the preceding endpoint, you may also apply filters here. The corresponding query parameters work in the same way.
The correlation ID can then be used to delete the audit log related to a particular solver request. This can be done easily by including the
correlation_ids parameter. It takes an array of correlation IDs separated by a comma, e.g. ., correlation_ids=af4012bd-d492-92ec-ffa4-31fd2b70b1bc,197d5d5a-f6f7-35de-1afb-dc26237ebfc9.If set, the
rules parameter limits the audit logs to be deleted to those produced by the specified rules. The individual rules are identified by ids separated by comma, i.e., rules=af4012bd-d492-92ec-ffa4-31fd2b70b1bc,197d5d5a-f6f7-35de-1afb-dc26237ebfc9. Moreover, one may further specify the allowed versions for each of the rules by including square brackets with the comma separated list of versions. For instance, rules=af4012bd-d492-92ec-ffa4-31fd2b70b1bc[1,2],197d5d5a-f6f7-35de-1afb-dc26237ebfc9 will return audits for rules with id af4012bd-d492-92ec-ffa4-31fd2b70b1bc whose version is either 1 or 2 and audits for rules with id 197d5d5a-f6f7-35de-1afb-dc26237ebfc9 of any version.It is also possible to filter solemnly by rule versions. To do that, use the square bracket expression without specifying rule IDs. For instance,
rules=[1,2] will return audit logs for rules whose version is 1 or 2, no matter the rule ID.The
date_gte and date_lte parameters allow to filter audit logs to be deleted by date (and time). The former serves as a lower bound on the date of the solve, the latter serves as an upper bound. In other words, audit logs will match this filter (and will be deleted) only if they are older than date_gte and younger than date_lte. When specifying one or both of the parameters, you can choose from a number of supported formats, including ISO 8601 and RFC 2822 Date time. (For some examples, see the preceding endpoint.)If set, the
status_codes parameter limits the audit logs to those produced with the specified status code. The individual codes are separated by comma. For example, to delete logs from rule solves that returned OK, use status_codes=200. To delete audit logs from rule solves that returned some kind of rule error, use status_codes=400, 401, 404, 406, 426. Eventually, to delete audit logs from rule solves that returned server error, use status_codes=500.Last modified 6mo ago