You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
title: 'Diagnose, troubleshoot, and solve issues - Azure Time Series Insights | Microsoft Docs'
2
+
title: 'Diagnose, troubleshoot, and solve issues - Azure Time Series Insights'
3
3
description: This article describes how to diagnose, troubleshoot, and solve common issues in your Azure Time Series Insights environment.
4
4
ms.service: time-series-insights
5
5
services: time-series-insights
6
6
author: deepakpalled
7
7
ms.author: dpalled
8
8
manager: cshankar
9
-
ms.reviewer: v-mamcge, jasonh, kfile
9
+
ms.reviewer: v-mamcge
10
10
ms.workload: big-data
11
11
ms.topic: troubleshooting
12
12
ms.date: 02/04/2020
@@ -15,121 +15,120 @@ ms.custom: seodec18
15
15
16
16
# Diagnose and solve issues in your Time Series Insights environment
17
17
18
-
This article describes some issues you might encounter in your Azure Time Series Insights environment. The article offers potential causes and solutions for resolution.
18
+
This article describes issues that you might encounter in your Azure Time Series Insights environment. The article offers potential causes and solutions for resolution.
19
19
20
20
## Video
21
21
22
-
### Learn about common Time Series Insights customer challenges and mitigations.</br>
22
+
### Learn about common Time Series Insights challenges and mitigations</br>
No data in the [Azure Time Series Insights explorer](https://insights.timeseries.azure.com) might occur for several common reasons:
28
+
If no data is appearing in the [Azure Time Series Insights explorer](https://insights.timeseries.azure.com), consider these common causes.
29
29
30
-
### Cause A: event source data isn't in JSON format
30
+
### Cause A: Event source data isn't in JSON format
31
31
32
-
Azure Time Series Insights only supports JSON data. For JSON samples, read[Supported JSON shapes](./how-to-shape-query-json.md).
32
+
Azure Time Series Insights supports only JSON data. For JSON samples, see[Supported JSON shapes](./how-to-shape-query-json.md).
33
33
34
-
### Cause B: the event source key is missing a required permission
34
+
### Cause B: The event source key is missing a required permission
35
35
36
-
* For an IoT hub in Azure IoT Hub, you must provide the key that has **service connect** permissions. Select either the **iothubowner** or **service**policies since both have **service connect** permissions.
36
+
* For an IoT hub in Azure IoT Hub, you need to provide the key that has service connect permissions. Select either the **iothubowner** or **service**policy. Both have service connect permissions.
37
37
38
38
[](media/diagnose-and-solve-problems/iothub-serviceconnect-permissions.png#lightbox)
39
39
40
-
* For an event hub in Azure Event Hubs, you must provide the key that has **listen** permissions. Either of the **read**or**manage** policies will work because they both have **listen** permissions.
40
+
* For an event hub in Azure Event Hubs, you need to provide the key that has listen permissions. The **read**and**manage** policies will both work because they both have listen permissions.
### Cause C: the consumer group provided isn't exclusive to Time Series Insights
44
+
### Cause C: The provided consumer group isn't exclusive to Time Series Insights
45
45
46
46
When you register an IoT hub or an event hub, it's important to set the consumer group that you want to use to read the data. This consumer group *can't be shared*. If the consumer group is shared, the underlying IoT hub or event hub automatically and randomly disconnects one of the readers. Provide a unique consumer group for Time Series Insights to read from.
47
47
48
-
### Cause D: the environment has just been provisioned
48
+
### Cause D: The environment has just been provisioned
49
49
50
50
Data will appear in your Time Series Insights explorer within a few minutes after the environment and its data are first created.
51
51
52
-
## Problem: some data is shown, but data is missing
52
+
## Problem: Some data is shown, but data is missing
53
53
54
-
When data appears only partially and the data appears to be lagging, you should consider several possibilities.
54
+
When data appears only partially and the data seems to be lagging, consider these possible issues.
55
55
56
-
### Cause A: your environment is being throttled
56
+
### Cause A: Your environment is being throttled
57
57
58
-
[Throttling](time-series-insights-environment-mitigate-latency.md) is a common issue when environments are provisioned after you create an event source that has data. Azure IoT Hub and Azure Events Hubs store data for up to seven days. Time Series Insights always start with the oldest event in the event source (first-in, first-out, or *FIFO*).
58
+
[Throttling](time-series-insights-environment-mitigate-latency.md) is a common issue when environments are provisioned after you create an event source that has data. Azure IoT Hub and Azure Events Hubs store data for up to seven days. Time Series Insights always starts with the oldest event in the event source (first-in, first-out, or *FIFO*).
59
59
60
-
For example, if you have 5 million events in an event source when you connect to an S1, single-unit Time Series Insights environment, Time Series Insights reads approximately 1 million events per day. It might look like Time Series Insights is experiencing five days of latency. However, what's happening is that the environment is being throttled.
60
+
For example, if you have 5 million events in an event source when you connect to an S1, single-unit Time Series Insights environment, Time Series Insights reads approximately 1 million events per day. It might look like Time Series Insights is experiencing five days of latency. But what's happening is that the environment is being throttled.
61
61
62
62
If you have old events in your event source, you can approach throttling in one of two ways:
63
63
64
64
- Change your event source's retention limits to help remove old events that you don't want to show up in Time Series Insights.
65
-
- Provision a larger environment size (number of units) to increase the throughput of old events. Using the preceding example, if you increase the same S1 environment to five units for one day, the environment should catch up within a day. If your steady-state event production is 1 million or fewer events per day, you can reduce the capacity of the event to one unit after it catches up.
65
+
- Provision a larger environment size (number of units) to increase the throughput of old events. In the preceding example, if you increase the same S1 environment to five units for one day, the environment should catch up within a day. If your steady-state event production is 1 million or fewer events per day, you can reduce the event capacity to one unit after Time Series Insights catches up.
66
66
67
-
The throttling limit is enforced based on the environment's SKU type and capacity. All event sources in the environment share this capacity. If the event source for your IoT hub or event hub pushes data beyond the enforced limits, you'll experience throttling and a lag.
67
+
The enforced throttling limit is based on the environment's SKU type and capacity. All event sources in the environment share this capacity. If the event source for your IoT hub or event hub pushes data beyond the enforced limits, you'll experience throttling and a lag.
68
68
69
69
The following figure shows a Time Series Insights environment that has an SKU of S1 and a capacity of 3. It can ingress 3 million events per day.
70
70
71
-
[](media/diagnose-and-solve-problems/environment-sku-current-capacity.png#lightbox)
As an example, assume an environment ingests messages from an event hub. The daily ingress rate is ~67,000 messages. This rate translates to approximately 46 messages every minute.
73
+
Imagine an environment that ingests messages from an event hub. It has a daily ingress rate of about 67,000 messages. This rate translates to approximately 46 messages every minute.
74
74
75
-
* If each event hub message is flattened to a single Time Series Insights event, throttling doesn't occur.
76
-
* If each event hub message is flattened to 100 Time Series Insights events, 4,600 events should be ingested every minute.
75
+
* If each event hub message is flattened to a single Time Series Insights event, throttling doesn't occur.
76
+
* If each event hub message is flattened to 100 Time Series Insights events, 4,600 events should be ingested every minute.
77
77
78
-
An S1 SKU environment that has a capacity of 3 can ingress only 2,100 events every minute (1 million events per day = 700 events per minute at three units = 2,100 events per minute).
78
+
An S1 SKU environment that has a capacity of 3 can ingress only 2,100 events every minute (1 million events per day = 700 events per minute at three units = 2,100 events per minute).
79
79
80
-
For a high-level understanding of how flattening logic works, read[Supported JSON shapes](./how-to-shape-query-json.md).
80
+
To get a high-level understanding of how flattening logic works, see[Supported JSON shapes](./how-to-shape-query-json.md).
81
81
82
82
#### Recommended resolutions for excessive throttling
83
83
84
84
To fix the lag, increase the SKU capacity of your environment. For more information, read [Scale your Time Series Insights environment](time-series-insights-how-to-scale-your-environment.md).
85
85
86
-
### Cause B: initial ingestion of historical data slows ingress
86
+
### Cause B: Initial ingestion of historical data slows ingress
87
87
88
88
If you connect an existing event source, it's likely that your IoT hub or event hub already contains data. The environment starts pulling data from the beginning of the event source's message retention period. This default processing can't be overridden. You can engage throttling. Throttling might take a while to catch up as it ingests historical data.
89
89
90
90
#### Recommended resolutions for large initial ingestion
91
91
92
92
To fix the lag:
93
93
94
-
1. Increase the SKU capacity to the maximum allowed value (10, in this case). After you increase capacity, the ingress process starts to catch up much more quickly. You are charged for the increased capacity. To visualize how quickly you're catching up, you can view the availability chart in the [Time Series Insights explorer](https://insights.timeseries.azure.com).
94
+
1. Increase the SKU capacity to the maximum allowed value (10, in this case). After you increase capacity, the ingress process starts to catch up much more quickly. You're charged for the increased capacity. To visualize how quickly you're catching up, you can view the availability chart in the [Time Series Insights explorer](https://insights.timeseries.azure.com).
95
95
96
96
2. When the lag is caught up, decrease the SKU capacity to your normal ingress rate.
97
97
98
-
## Problem: data was showing previously but is no longer showing
98
+
## Problem: Data was showing previously but is no longer showing
99
99
100
-
TSI is no longer ingesting data but events are still streaming into Iot Hub or Event Hub
100
+
If Time Series Insights is no longer ingesting data, but events are still streaming into Iot Hub or Event Hub, consider this potential cause.
101
101
102
-
### Cause A: your hub access key was regenerated and your environment needs updating
102
+
### Cause A: Your hub access key was regenerated and your environment needs to be updated
103
103
104
-
This problem occurs when the key provided when creating your event source is no longer valid. You would see telemetry in your hub but no Ingress Received Messages in Time Series Insights. If you are unsure whether or not the key was regenerated you can search your Event Hubs' Activity log for "Create or Update Namespace Authorization Rules" or search "Create or update IotHub Resource" for IoT hub.
104
+
This problem occurs when the key provided when you created your event source is no longer valid. You would see telemetry in your hub but no Ingress Received Messages in Time Series Insights. If you aren't sure whether the key was regenerated, you can search your event hub's activity log for "Create or Update Namespace Authorization Rules." For an IoT hub, search for "Create or update IotHub Resource."
105
105
106
-
To update your Time Series Insights environment with the new key open your hub resource in the Azure portal and copy the new key. Navigate to your TSI resource and click on Event Sources.
106
+
To update your Time Series Insights environment with the new key, open your hub resource in the Azure portal and copy the new key. Go to your Time Series Insights resource and select **Event Sources**:
[](media/diagnose-and-solve-problems/update-hub-key-step-2.png#lightbox)
113
113
114
-
## Problem: my event source's timestamp property name setting doesn't work
114
+
## Problem: The event source's timestamp property name setting doesn't work
115
115
116
-
Ensure that the timestamp property name and value conform to the following rules:
116
+
Ensure that the timestamp property value that comes from your event source as a JSON string is in the format _yyyy-MM-ddTHH:mm:ss.FFFFFFFK_. Here's an example: **2008-04-12T12:53Z**.
117
117
118
-
* The timestamp property name is case-sensitive.
119
-
* The timestamp property value that comes from your event source as a JSON string should have the format _yyyy-MM-ddTHH:mm:ss.FFFFFFFK_. An example is **2008-04-12T12:53Z**.
118
+
Keep in mind that the timestamp property name is case-sensitive.
120
119
121
-
The easiest way to ensure that your timestamp property name is captured and working properly is to use the Time Series Insights explorer. In the Time Series Insights explorer, using the chart, select a period of time after you entered the timestamp property name. Right-click the selection, and then select the **Explore events** option.
120
+
The easiest way to ensure that your timestamp property name is captured and working properly is to use the Time Series Insights explorer. In the Time Series Insights explorer, using the chart, select a period of time after you entered the timestamp property name. Right-click the selection, and then select **Explore events**.
122
121
123
122
The first column header should be your timestamp property name. Next to the word **Timestamp**, **($ts)** will be displayed.
124
123
125
-
The following values will not be displayed:
124
+
The following values won't be displayed:
126
125
127
126
-*(abc)*: Indicates that Time Series Insights is reading the data values as strings.
128
-
-*Calendar icon*: Indicates that Time Series Insights is reading the data value as *datetime*.
129
-
-*#*: Indicates that Time Series Insights is reading the data values as an integer.
127
+
-*Calendar icon*: Indicates that Time Series Insights is reading the data values as datetime values.
128
+
-*#*: Indicates that Time Series Insights is reading the data values as integers.
130
129
131
130
## Next steps
132
131
133
-
- Read about [How to mitigate latency in Azure Time Series Insights](time-series-insights-environment-mitigate-latency.md).
132
+
- Read about [how to mitigate latency in Azure Time Series Insights](time-series-insights-environment-mitigate-latency.md).
134
133
135
-
- Learn [how to scale your Time Series Insights environment](time-series-insights-how-to-scale-your-environment.md).
134
+
- Learn [how to scale your Time Series Insights environment](time-series-insights-how-to-scale-your-environment.md).
0 commit comments