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
[ON week] Optimize AutoOps section for gen AI (#3964)
This PR updates and restructures some of the AutoOps docs to optimize
them for gen AI retrieval. This includes applying general GEO strategies
like structural consistency, navigational context, chunking, etc. +
making targeted changes to address some AI hallucinations.
Copy file name to clipboardExpand all lines: deploy-manage/monitor/autoops.md
+2-1Lines changed: 2 additions & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -19,7 +19,8 @@ products:
19
19
AutoOps diagnoses issues in {{es}} by analyzing hundreds of metrics, providing root-cause analysis and accurate resolution paths. With AutoOps, customers can prevent and resolve issues, cut down administration time, and optimize resource utilization.
Copy file name to clipboardExpand all lines: deploy-manage/monitor/autoops/cc-cloud-connect-autoops-troubleshooting.md
+1-1Lines changed: 1 addition & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -38,7 +38,7 @@ $$$next-steps$$$**After running the installation command, I can't move on to the
38
38
: If an error appears on the screen, follow the suggestion in the error message and try to run the command again. If the issue is not resolved, explore [additional resources](/troubleshoot/index.md#troubleshoot-additional-resources) or [contact us](/troubleshoot/index.md#contact-us).
39
39
40
40
$$$firewall$$$**My organization's firewall may be preventing {{agent}} from collecting and sending metrics.**
41
-
: If yoususpect that a corporate firewall is hindering {{agent}}, complete the following steps to test each component of the connection and implement an appropriate solution.
41
+
: If you're having issues with connecting your self-managed cluster to AutoOps and you suspect that a firewall may be the reason, complete the following steps to test each component of the connection and implement an appropriate solution.
42
42
43
43
:::{tip}
44
44
Run the following tests within the context of your execution environment. That is, if your chosen installation method is Kubernetes, run the commands from within the pod; for Docker, run the commands from within the container, and so on.
Copy file name to clipboardExpand all lines: deploy-manage/monitor/autoops/cc-connect-self-managed-to-autoops.md
+4Lines changed: 4 additions & 0 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -102,6 +102,10 @@ Select one of the following methods to install {{agent}}:
102
102
* **Windows**
103
103
-->
104
104
105
+
:::{note}
106
+
macOS is not a supported platform for installing {{agent}} to connect to AutoOps in a self-managed production environment. However, you can use macOS to [connect your local development cluster to AutoOps](/deploy-manage/monitor/autoops/cc-connect-local-dev-to-autoops.md).
107
+
:::
108
+
105
109
:::{important}
106
110
Using AutoOps for your ECE, ECK, and self-managed clusters requires a new, dedicated {{agent}}. You must install an agent even if you already have an existing one for other purposes.
Copy file name to clipboardExpand all lines: deploy-manage/monitor/autoops/ec-autoops-deployment-view.md
+25-17Lines changed: 25 additions & 17 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -16,43 +16,51 @@ products:
16
16
17
17
# Deployment or Cluster view in AutoOps [ec-autoops-deployment-view]
18
18
19
-
The **Deployment** view (for {{ECH}} deployments) or **Cluster** view (for ECE, ECK, and self-managed clusters), is the event control panel that allows you to see which issues are affecting the {{es}} cluster and get a list of action items to address them.
19
+
The **Deployment** view (for {{ECH}} deployments) or **Cluster** view (for ECE, ECK, and self-managed clusters), is the event control panel that gives you an overview of the events, resource usage, and performance of your deployments or clusters.
20
20
21
-
## Events over time [ec-autoops-events-over-time]
21
+
To get to this view, go to AutoOps in your deployment or cluster and select **Deployment** or **Cluster** from the side navigation.
22
22
23
-
The **Events Over Time** panel lists all the recent events the {{es}} cluster has triggered, ordered by criticality. It also gives a color-coded heat map to help understand when and how often a particular event happened. Click on any mosaic to get details about a particular event, for example the specific node/index/shard affected, event time and duration, and a detailed description of the actions you can take to mitigate that event.
23
+
You can use the **Deployment** or **Cluster** dropdown at the top of the screen to select which deployment or cluster you want to monitor.
24
24
25
-
Refer to [AutoOps events](ec-autoops-events.md) for more details.
25
+
## Panels in the Deployment or Cluster view
26
26
27
+
The **Deployment** or **Cluster** view shows the following panels.
27
28
28
-
##Open events [ec-autoops-open-events]
29
+
### Events over time [ec-autoops-events-over-time]
29
30
30
-
The **Open Events** panel lists open events sorted by severity and time. When the conditions that triggered the event no longer exist, the event is automatically set to close and appear in the **Events History** panel. Closing an event does not necessarily indicate that the customer resolved the issue, but rather that AutoOps no longer detects it.
31
+
The **Events Over Time** panel lists all the recent events the {{es}} cluster has triggered, ordered by criticality. It also displays a color-coded heat map to help understand when and how often a particular event happened. Click on any mosaic to get details about a particular event, such as the specific node, index, or shard affected, event time and duration, and a detailed description of the actions you can take to mitigate that event.
31
32
33
+
Refer to [AutoOps events](ec-autoops-events.md) for more details.
32
34
33
-
##Events History [ec-events-history]
35
+
### Open Events [ec-autoops-open-events]
34
36
35
-
The **Events History** panel lists events that happened at some point and that have been triggered, but some conditions changed and are no longer active. For example, when your cluster experiences a peak in search rate, that might trigger a "Too many tasks on queue" event. Now, your cluster is more relaxed in terms of search rate, so this event is no longer an issue, but it was recorded for historical reasons. Events history is also sorted by severity first and then by time.
37
+
The **Open Events** panel lists open events sorted by severity and time. When the conditions that triggered the event no longer exist, the event is automatically set to close and appear in the **Events History** panel. Closing an event does not necessarily indicate that the customer resolved the issue, but rather that AutoOps no longer detects it.
36
38
39
+
### Events History [ec-events-history]
37
40
38
-
## Resources [ec-deployment-resources]
41
+
The **Events History** panel lists events that were triggered in the past but are no longer active because of changed conditions.
42
+
43
+
Let's say your cluster experienced a peak in search rate, triggering a "Too many tasks on queue" event. Now, your cluster is more relaxed in terms of search rate, so this event is no longer an issue, but it was recorded for historical reasons. Like Open Events, Events History is also sorted first by severity and then by time.
44
+
45
+
### Resources [ec-deployment-resources]
39
46
40
47
The **Resources** panel provides a quick overview of {{es}} cluster resource usage. The resources are presented based on their respective data tiers and include JVM memory usage, CPU usage, and storage usage over time. You can view essential cluster information such as the {{es}} version, total number of nodes, total number of shards, and total volume of used storage.
The **Performance** panel shows the following key performance metrics aggregated at both the cluster level and the selected tier levels:
54
57
55
58
***Search rate**: The number of search requests executed per second across all shards in the deployment or cluster, as well as within the selected data tiers.
56
59
***Search latency**: The average latency of search operations across all shards in the deployment or cluster, and within the selected data tiers.
57
60
***Indexing rate**: The number of documents indexed per second across all shards in the deployment or cluster, as well as within the selected data tiers.
58
61
***Indexing latency**: The average latency of indexing operations across all shards in the deployment or cluster, and within the selected data tiers.
@@ -38,8 +39,7 @@ Whether you are using AutoOps in your [{{ech}} deployment](/deploy-manage/monito
38
39
39
40
**Setting up AutoOps for self-managed clusters**
40
41
*[Can I use Cloud Connect to connect my {{ech}} clusters to AutoOps?](#cc-autoops-ech)
41
-
*[Can I use AutoOps for my clusters if my environment is air-gapped?](#autoops-air-gapped)
42
-
*[Can I use macOS to install {{agent}} for this feature?](#macos-install)
42
+
*[Can I use macOS to set up AutoOps for my self-managed clusters?](#macos-install)
43
43
*[Do I have to define an Elastic IP address to enable the agent to send data to {{ecloud}}?](#elastic-ip-address)
44
44
45
45
**Collected metrics and data in AutoOps for self-managed clusters**
@@ -59,6 +59,9 @@ $$$autoops-deployment-types$$$**Where is AutoOps available?**
59
59
$$$cant-see-autoops$$$**Why can't I see AutoOps in some deployments and projects?**
60
60
AutoOps is rolling out in phases across CSPs and [regions](/deploy-manage/monitor/autoops/ec-autoops-regions.md), so you may not see it if your deployment or project is in a region where AutoOps is not available yet. AutoOps is currently not available in Azure and GCP.
61
61
62
+
$$$autoops-air-gapped$$$**Can I use AutoOps if my environment is air-gapped?**
63
+
: Not at this time. AutoOps is currently only available as a cloud service and you need an internet connection to send metrics to {{ecloud}}. For air-gapped environments, we plan to offer a locally deployable version in the future.
64
+
62
65
$$$autoops-license$$$**How is AutoOps licensed?**
63
66
: For {{ech}} deployments and {{serverless-short}} projects, AutoOps is available to {{ecloud}} customers at all subscription levels at no additional cost, and it does not consume ECU.
$$$autoops-air-gapped$$$**Can I use AutoOps for my clusters if my environment is air-gapped?**
106
-
: Not at this time. AutoOps is currently only available as a cloud service and you need an internet connection to send metrics to {{ecloud}}. For air-gapped environments, we plan to offer a locally deployable version in the future.
107
-
108
-
$$$macos-install$$$**Can I use macOS to install {{agent}} for this feature?**
109
-
: macOS is not a supported platform for installing {{agent}} and connecting your clusters to AutoOps.
108
+
$$$macos-install$$$**Can I use macOS to set up AutoOps for my self-managed clusters?**
109
+
: No, macOS is not a supported platform for setting up and connecting to AutoOps in a self-managed production environment. However, you can use macOS to [connect your local development cluster to AutoOps](/deploy-manage/monitor/autoops/cc-connect-local-dev-to-autoops.md).
110
110
111
111
$$$elastic-ip-address$$$**Do I have to define an Elastic IP address to enable the agent to send data to {{ecloud}}?**
112
112
: You may need to define an IP address if your organization’s settings will block the agent from sending out data.
Copy file name to clipboardExpand all lines: deploy-manage/monitor/autoops/ec-autoops-index-view.md
+10-20Lines changed: 10 additions & 20 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -16,32 +16,22 @@ products:
16
16
17
17
# Indices view in AutoOps [ec-autoops-index-view]
18
18
19
-
The **Indices** view provides detailed statistics for each {{es}} index in your deployment.
19
+
The **Indices** view provides detailed statistics for each {{es}} index in your deployment, allowing you to visualize index performance and trends to detect anomalies and optimize search efficiency.
The **Indices** view is essential for monitoring {{es}} indices, and offers comprehensive insights at a glance by displaying a clear and informative table about the following metrics:
21
+
Information about each index is organized into a clear table with columns for the index's name, primary and total shards, indexing and search rate per second, and more. You can expand each index entry to dive deeper into real-time metrics.
26
22
27
-
* Index Name
28
-
* Primary Shards and Total Shards
29
-
* Shard Size
30
-
* Size in Bytes
31
-
* Doc Count
32
-
* Indexing Rate/Sec
33
-
* Search Rate/Sec
34
-
* Index Latency
35
-
* Search Latency
23
+
To get to the **Indices** view, go to AutoOps in your deployment or cluster and select **Indices** from the side navigation.
36
24
37
-
You can expand each index entry to dive deeper into real-time metrics. This is an intuitive and dynamic feature that allows you to visualize index activities and trends, in order to detect anomalies and optimize search efficiency.
38
-
39
-
The **Indices** view offers you a powerful tool for managing and optimizing your {{es}} indices. By providing a detailed and up-to-date overview of index performance and usage, AutoOps ensures that your search and indexing operations run smoothly and efficiently.
Copy file name to clipboardExpand all lines: deploy-manage/monitor/autoops/ec-autoops-nodes-view.md
+19-4Lines changed: 19 additions & 4 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -16,15 +16,30 @@ products:
16
16
17
17
# Nodes view in AutoOps [ec-autoops-nodes-view]
18
18
19
-
The **Nodes** view provides a thorough overview on the essential metrics for all monitored nodes. You can delve into specific nodes to observe metrics over extended periods. This includes data on the indexing rate and latency, search rate and latency, as well as details concerning thread pools, data, circuit breakers, network, disk, and additional elements.
19
+
The **Nodes** view provides a thorough look into essential metrics for all monitored nodes. With this view, you can gain deeper insight into your cluster's health on a per-node basis and observe each metric over extended periods of time.
20
+
21
+
To get to the **Nodes** view, go to AutoOps in your deployment or cluster and select **Nodes** from the side navigation.
:alt: Screenshot showing the Nodes view in AutoOps
23
26
:::
24
27
25
-
Similar to the **Deployment** or **Cluster** view, the list of open events is sorted by severity and open time.
28
+
## Panels in the Nodes view
29
+
30
+
The **Nodes** view shows the following panels.
31
+
32
+
### Nodes
33
+
34
+
The **Nodes** table lists all the nodes used by the {{es}} cluster, along with their name, role, and status. The elected master node is marked with a star.
35
+
36
+
### Open Events
37
+
38
+
The **Open Events** panel lists open events sorted by severity and time. When the conditions that triggered the event no longer exist, the event is automatically set to close and appear in the **Events History** panel. Closing an event does not necessarily indicate that the customer resolved the issue, but rather that AutoOps no longer detects it.
39
+
40
+
### Panels covering other monitoring areas
26
41
27
-
The following table lists all the nodes used by the {{es}} cluster, presenting node name, role and status. The elected master node will be marked with a start sign.
42
+
The following table lists all the other panels in the **Nodes** view that drill down into specific monitoring areas, along with the names and descriptions of metrics they present.
0 commit comments