Skip to content

Commit 861fd81

Browse files
tiburdtiburd
authored
Merge pull request #101162 from ronortloff/master
Update sql-data-warehouse-workload-management-portal-monitor.md
2 parents a8bf39b + 57d6fb1 commit 861fd81

File tree

2 files changed

+30
-28
lines changed

2 files changed

+30
-28
lines changed

‎articles/sql-data-warehouse/sql-data-warehouse-concept-resource-utilization-query-activity.md

Lines changed: 20 additions & 18 deletions
Original file line numberDiff line numberDiff line change
@@ -7,7 +7,7 @@ manager: craigg-msft
77
ms.service: sql-data-warehouse
88
ms.topic: conceptual
99
ms.subservice: manage
10-
ms.date: 08/09/2019
10+
ms.date: 01/14/2020
1111
ms.author: kevin
1212
ms.reviewer: igorstan
1313
ms.custom: seo-lt-2019
@@ -22,23 +22,25 @@ The following metrics are available in the Azure portal for SQL Data Warehouse.
2222

2323
| Metric Name | Description | Aggregation Type |
2424
| ----------------------- | ------------------------------------------------------------ | ---------------- |
25-
| CPU percentage | CPU utilization across all nodes for the data warehouse | Maximum |
26-
| Data IO percentage | IO Utilization across all nodes for the data warehouse | Maximum |
27-
| Memory percentage | Memory utilization (SQL Server) across all nodes for the data warehouse | Maximum |
28-
| Successful Connections | Number of successful connections to the data | Total |
29-
| Failed Connections | Number of failed connections to the data warehouse | Total |
30-
| Blocked by Firewall | Number of logins to the data warehouse which was blocked | Total |
31-
| DWU limit | Service level objective of the data warehouse | Maximum |
32-
| DWU percentage | Maximum between CPU percentage and Data IO percentage | Maximum |
33-
| DWU used | DWU limit * DWU percentage | Maximum |
34-
| Cache hit percentage | (cache hits / cache miss) * 100 where cache hits is the sum of all columnstore segments hits in the local SSD cache and cache miss is the columnstore segments misses in the local SSD cache summed across all nodes | Maximum |
35-
| Cache used percentage | (cache used / cache capacity) * 100 where cache used is the sum of all bytes in the local SSD cache across all nodes and cache capacity is the sum of the storage capacity of the local SSD cache across all nodes | Maximum |
36-
| Local tempdb percentage | Local tempdb utilization across all compute nodes - values are emitted every five minutes | Maximum |
37-
38-
> Things to consider when viewing metrics and setting alerts:
39-
>
40-
> - Failed and successful connections are reported for a particular data warehouse - not for the logical server
41-
> - Memory percentage reflects utilization even if the data warehouse is in idle state - it does not reflect active workload memory consumption. Use and track this metric along with others (tempdb, gen2 cache) to make a holistic decision on if scaling for additional cache capacity will increase workload performance to meet your requirements.
25+
| CPU percentage | CPU utilization across all nodes for the data warehouse | Avg, Min, Max |
26+
| Data IO percentage | IO Utilization across all nodes for the data warehouse | Avg, Min, Max |
27+
| Memory percentage | Memory utilization (SQL Server) across all nodes for the data warehouse | Avg, Min, Max |
28+
| Active Queries | Number of active queries executing on the system | Sum |
29+
| Queued Queries | Number of queued queries waiting to start executing | Sum |
30+
| Successful Connections | Number of successful connections to the data | Sum, Count |
31+
| Failed Connections | Number of failed connections to the data warehouse | Sum, Count |
32+
| Blocked by Firewall | Number of logins to the data warehouse which was blocked | Sum, Count |
33+
| DWU limit | Service level objective of the data warehouse | Avg, Min, Max |
34+
| DWU percentage | Maximum between CPU percentage and Data IO percentage | Avg, Min, Max |
35+
| DWU used | DWU limit * DWU percentage | Avg, Min, Max |
36+
| Cache hit percentage | (cache hits / cache miss) * 100 where cache hits is the sum of all columnstore segments hits in the local SSD cache and cache miss is the columnstore segments misses in the local SSD cache summed across all nodes | Avg, Min, Max |
37+
| Cache used percentage | (cache used / cache capacity) * 100 where cache used is the sum of all bytes in the local SSD cache across all nodes and cache capacity is the sum of the storage capacity of the local SSD cache across all nodes | Avg, Min, Max |
38+
| Local tempdb percentage | Local tempdb utilization across all compute nodes - values are emitted every five minutes | Avg, Min, Max |
39+
40+
Things to consider when viewing metrics and setting alerts:
41+
42+
- Failed and successful connections are reported for a particular data warehouse - not for the logical server
43+
- Memory percentage reflects utilization even if the data warehouse is in idle state - it does not reflect active workload memory consumption. Use and track this metric along with others (tempdb, gen2 cache) to make a holistic decision on if scaling for additional cache capacity will increase workload performance to meet your requirements.
4244

4345

4446
## Query activity

‎articles/sql-data-warehouse/sql-data-warehouse-workload-management-portal-monitor.md

Lines changed: 10 additions & 10 deletions
Original file line numberDiff line numberDiff line change
@@ -7,7 +7,7 @@ manager: craigg
77
ms.service: sql-data-warehouse
88
ms.topic: conceptual
99
ms.subservice: workload-management
10-
ms.date: 01/13/2020
10+
ms.date: 01/14/2020
1111
ms.author: rortloff
1212
ms.reviewer: jrasnick
1313
ms.custom: seo-lt-2019
@@ -45,10 +45,10 @@ CREATE WORKLOAD CLASSIFIER wcCEOPriority
4545
WITH ( WORKLOAD_GROUP = 'wgPriority'
4646
,MEMBERNAME = 'TheCEO');
4747
```
48-
The below chart is configured as follows:
49-
Metric 1: *Effective min resource percent* (Avg aggregation, `blue line`)
50-
Metric 2: *Workload group allocation by system percent* (Avg aggregation, `purple line`)
51-
Filter: [Workload Group] = `wgPriority`
48+
The below chart is configured as follows:<br>
49+
Metric 1: *Effective min resource percent* (Avg aggregation, `blue line`)<br>
50+
Metric 2: *Workload group allocation by system percent* (Avg aggregation, `purple line`)<br>
51+
Filter: [Workload Group] = `wgPriority`<br>
5252
![underutilized-wg.png](media/sql-data-warehouse-workload-management-portal-monitor/underutilized-wg.png)
5353
The chart shows that with 25% workload isolation, only 10% is being used on average. In this case, the `MIN_PERCENTAGE_RESOURCE` parameter value could be lowered to between 10 or 15 and allow for other workloads on the system to consume the resources.
5454

@@ -65,11 +65,11 @@ CREATE WORKLOAD CLASSIFIER wcDataAnalyst
6565
WITH ( WORKLOAD_GROUP = 'wgDataAnalyst'
6666
,MEMBERNAME = 'DataAnalyst');
6767
```
68-
The below chart is configured as follows:
69-
Metric 1: *Effective cap resource percent* (Avg aggregation, `blue line`)
70-
Metric 2: *Workload group allocation by max resource percent* (Avg aggregation, `purple line`)
71-
Metric 3: *Workload group queued queries* (Sum aggregation, `turquoise line`)
72-
Filter: [Workload Group] = `wgDataAnalyst`
68+
The below chart is configured as follows:<br>
69+
Metric 1: *Effective cap resource percent* (Avg aggregation, `blue line`)<br>
70+
Metric 2: *Workload group allocation by max resource percent* (Avg aggregation, `purple line`)<br>
71+
Metric 3: *Workload group queued queries* (Sum aggregation, `turquoise line`)<br>
72+
Filter: [Workload Group] = `wgDataAnalyst`<br>
7373
![bottle-necked-wg](media/sql-data-warehouse-workload-management-portal-monitor/bottle-necked-wg.png)
7474
The chart shows that with a 9% cap on resources, the workload group is 90%+ utilized (from the *Workload group allocation by max resource percent metric*). There is a steady queuing of queries as shown from the *Workload group queued queries metric*. In this case, increasing the `CAP_PERCENTAGE_RESOURCE` to a value higher than 9% will allow more queries to execute concurrently. Increasing the `CAP_PERCENTAGE_RESOURCE` assumes that there are enough resources available and not isolated by other workload groups. Verify the cap increased by checking the *Effective cap resource percent metric*. If more throughput is desired, also consider increasing the `REQUEST_MIN_RESOURCE_GRANT_PERCENT` to a value greater than 3. Increasing the `REQUEST_MIN_RESOURCE_GRANT_PERCENT` could allow queries to run faster.
7575

0 commit comments

Comments
 (0)