Kusto Query Pipeline Query

Question

I am trying to create another column from a pipeline runs data using Azure data explorer/Kusto queries. I am very new to Kusto and not sure how to go about that. Goal is for each customer,

1. If previous run failed and last run failed get difference of days/hours between failures.
2. If previous run succeeded and last run failed, get difference of days/hours between the events
3. If previous event failed and last event succeeded, then ignore.

Dataset

Customers   PipelineType    PipelineState    TimeStamp
CustomerA   PipelineA   Succes               2021-08-13 12:59:03.0073653
CustomerA   PipelineA   Fail                 2021-08-13 09:59:03.0124853
CustomerA   PipelineB   Succes               2021-08-13 11:56:03.0151948
CustomerA   Pipeline B  Fail                 2021-08-12 17:56:03.0019445
CustomerA   Pipeline C  Succes               2021-08-13 13:16:03.0015617
CustomerA   Pipeline C  Fail                 2021-07-30 21:52:03.0157372
CustomerB   PipelineA   Succes               2021-08-13 12:59:03.0073331
CustomerB   PipelineA   Succes               2021-08-13 12:57:03.0099138
CustomerB   PipelineB   Fail                 2021-07-30 03:33:03.0123262
CustomerB   Pipeline B  Succes               2021-08-13 13:16:03.0015297
CustomerB   Pipeline C  Fail                 2021-08-13 12:57:03.0099499
CustomerB   Pipeline C  Succes               2021-08-13 13:16:03.0016348
CustomerC   PipelineA   Succes               2021-08-13 13:16:03.0016999
CustomerC   PipelineA   Succes               2021-08-13 12:59:03.0074113
CustomerC   PipelineB   Succes               2021-08-13 10:56:03.0075546
CustomerC   Pipeline B  Fail                 2021-08-11 06:54:03.0118628
CustomerC   Pipeline C  Fail                 2021-08-13 13:16:03.0016233
CustomerC   Pipeline C  Fail                 2021-08-13 12:59:03.0072337
``

Answer 1

If I understand the requirements correctly, you could sort your data set and then use the case() and prev() functions.

For example:

datatable(customer:string, PipelineType:string, PipelineState:string, TimeStamp:datetime)
[
    'CustomerA', 'Pipeline A', 'Fail',    datetime(2021-08-13 12:59:03.0073653),
    'CustomerA', 'Pipeline A', 'Fail',    datetime(2021-08-13 09:59:03.0124853),
    'CustomerA', 'Pipeline B', 'Success', datetime(2021-08-13 11:56:03.0151948),
    'CustomerA', 'Pipeline B', 'Fail',    datetime(2021-08-12 17:56:03.0019445),
    'CustomerA', 'Pipeline C', 'Success', datetime(2021-08-13 13:16:03.0015617),
    'CustomerA', 'Pipeline C', 'Fail',    datetime(2021-07-30 21:52:03.0157372),
    'CustomerB', 'Pipeline A', 'Fail',    datetime(2021-08-13 12:59:03.0073331),
    'CustomerB', 'Pipeline A', 'Success', datetime(2021-08-13 12:57:03.0099138),
    'CustomerB', 'Pipeline B', 'Fail',    datetime(2021-07-30 03:33:03.0123262),
    'CustomerB', 'Pipeline B', 'Success', datetime(2021-08-13 13:16:03.0015297),
    'CustomerB', 'Pipeline C', 'Fail',    datetime(2021-08-13 12:57:03.0099499),
    'CustomerB', 'Pipeline C', 'Success', datetime(2021-08-13 13:16:03.0016348),
    'CustomerC', 'Pipeline A', 'Fail',    datetime(2021-08-13 13:16:03.0016999),
    'CustomerC', 'Pipeline A', 'Fail',    datetime(2021-08-13 12:59:03.0074113),
    'CustomerC', 'Pipeline B', 'Success', datetime(2021-08-13 10:56:03.0075546),
    'CustomerC', 'Pipeline B', 'Fail',    datetime(2021-08-11 06:54:03.0118628),
    'CustomerC', 'Pipeline C', 'Fail',    datetime(2021-08-13 13:16:03.0016233),
    'CustomerC', 'Pipeline C', 'Fail',    datetime(2021-08-13 12:59:03.0072337),
]   
| order by customer asc, PipelineType asc, TimeStamp asc
| extend result = case(prev(customer) == customer and prev(PipelineType) == PipelineType and PipelineState == 'Fail', TimeStamp - prev(TimeStamp), timespan(null))

customer	PipelineType	PipelineState	TimeStamp	result
CustomerA	Pipeline A	Fail	2021-08-13 09:59:03.0124853
CustomerA	Pipeline A	Fail	2021-08-13 12:59:03.0073653	02:59:59.9948800
CustomerA	Pipeline B	Fail	2021-08-12 17:56:03.0019445
CustomerA	Pipeline B	Success	2021-08-13 11:56:03.0151948
CustomerA	Pipeline C	Fail	2021-07-30 21:52:03.0157372
CustomerA	Pipeline C	Success	2021-08-13 13:16:03.0015617
CustomerB	Pipeline A	Success	2021-08-13 12:57:03.0099138
CustomerB	Pipeline A	Fail	2021-08-13 12:59:03.0073331	00:01:59.9974193
CustomerB	Pipeline B	Fail	2021-07-30 03:33:03.0123262
CustomerB	Pipeline B	Success	2021-08-13 13:16:03.0015297
CustomerB	Pipeline C	Fail	2021-08-13 12:57:03.0099499
CustomerB	Pipeline C	Success	2021-08-13 13:16:03.0016348
CustomerC	Pipeline A	Fail	2021-08-13 12:59:03.0074113
CustomerC	Pipeline A	Fail	2021-08-13 13:16:03.0016999	00:16:59.9942886
CustomerC	Pipeline B	Fail	2021-08-11 06:54:03.0118628
CustomerC	Pipeline B	Success	2021-08-13 10:56:03.0075546
CustomerC	Pipeline C	Fail	2021-08-13 12:59:03.0072337
CustomerC	Pipeline C	Fail	2021-08-13 13:16:03.0016233	00:16:59.9943896

---

Update : in reply to your comment - just add the appropriate filters.

For example:

datatable(customer:string, PipelineType:string, PipelineState:string, TimeStamp:datetime)
[
    'CustomerA', 'Pipeline A', 'Fail',    datetime(2021-08-13 12:59:03.0073653),
    'CustomerA', 'Pipeline A', 'Fail',    datetime(2021-08-13 09:59:03.0124853),
    'CustomerA', 'Pipeline B', 'Success', datetime(2021-08-13 11:56:03.0151948),
    'CustomerA', 'Pipeline B', 'Fail',    datetime(2021-08-12 17:56:03.0019445),
    'CustomerA', 'Pipeline C', 'Success', datetime(2021-08-13 13:16:03.0015617),
    'CustomerA', 'Pipeline C', 'Fail',    datetime(2021-07-30 21:52:03.0157372),
    'CustomerB', 'Pipeline A', 'Fail',    datetime(2021-08-13 12:59:03.0073331),
    'CustomerB', 'Pipeline A', 'Success', datetime(2021-08-13 12:57:03.0099138),
    'CustomerB', 'Pipeline B', 'Fail',    datetime(2021-07-30 03:33:03.0123262),
    'CustomerB', 'Pipeline B', 'Success', datetime(2021-08-13 13:16:03.0015297),
    'CustomerB', 'Pipeline C', 'Fail',    datetime(2021-08-13 12:57:03.0099499),
    'CustomerB', 'Pipeline C', 'Success', datetime(2021-08-13 13:16:03.0016348),
    'CustomerC', 'Pipeline A', 'Fail',    datetime(2021-08-13 13:16:03.0016999),
    'CustomerC', 'Pipeline A', 'Fail',    datetime(2021-08-13 12:59:03.0074113),
    'CustomerC', 'Pipeline B', 'Success', datetime(2021-08-13 10:56:03.0075546),
    'CustomerC', 'Pipeline B', 'Fail',    datetime(2021-08-11 06:54:03.0118628),
    'CustomerC', 'Pipeline C', 'Fail',    datetime(2021-08-13 13:16:03.0016233),
    'CustomerC', 'Pipeline C', 'Fail',    datetime(2021-08-13 12:59:03.0072337),
]   
| order by customer asc, PipelineType asc, TimeStamp asc
| where not((prev(customer) == customer and prev(PipelineType) == PipelineType and PipelineState == 'Success' and prev(PipelineState) == 'Fail') or 
            (prev(customer) == customer and prev(PipelineType) == PipelineType and PipelineState == 'Fail' and next(PipelineState) == 'Success'))
| extend result = case(prev(customer) == customer and prev(PipelineType) == PipelineType and PipelineState == 'Fail', TimeStamp - prev(TimeStamp), timespan(null))

customer	PipelineType	PipelineState	TimeStamp	result
CustomerA	Pipeline A	Fail	2021-08-13 09:59:03.0124853
CustomerA	Pipeline A	Fail	2021-08-13 12:59:03.0073653	02:59:59.9948800
CustomerA	Pipeline B	Fail	2021-08-12 17:56:03.0019445
CustomerA	Pipeline C	Fail	2021-07-30 21:52:03.0157372
CustomerB	Pipeline A	Success	2021-08-13 12:57:03.0099138
CustomerB	Pipeline A	Fail	2021-08-13 12:59:03.0073331	00:01:59.9974193
CustomerB	Pipeline B	Fail	2021-07-30 03:33:03.0123262
CustomerB	Pipeline C	Fail	2021-08-13 12:57:03.0099499
CustomerC	Pipeline A	Fail	2021-08-13 12:59:03.0074113
CustomerC	Pipeline A	Fail	2021-08-13 13:16:03.0016999	00:16:59.9942886
CustomerC	Pipeline B	Fail	2021-08-11 06:54:03.0118628
CustomerC	Pipeline C	Fail	2021-08-13 12:59:03.0072337
CustomerC	Pipeline C	Fail	2021-08-13 13:16:03.0016233	00:16:59.9943896