Elasticsearch:View instance task progress

Limitations

The following features are not supported for Elasticsearch 7.16 instances or any instances in the China (Chengdu) and China (Guangzhou) regions:

• Viewing the stages, progress details, and change records of a change task.

• Viewing data migration progress.

Procedure

1. Log on to the Alibaba Cloud Elasticsearch console.

2. In the navigation pane on the left, click Elasticsearch Clusters.

3. Go to the target instance.

   1. In the top navigation bar, select a resource group and region.

   2. On the Elasticsearch Clusters page, click the ID of the target instance.

4. In the upper-right corner, click the icon to open the Tasks panel, where you can view the instance change progress and data migration progress.

   Instance change progress

   Click the Change Progress tab to view the progress details of the change task. After an instance change is triggered, the change task list is updated with a delay of about 10 seconds.

   The Change Progress tab contains the following parameters and actions:

   Parameter/action	Description
   Change Start Time	The time when the instance change started.
   Change Type	The type of instance change, such as creating an instance, restarting an instance, or changing the cluster configuration.
   Changed By	The ID of the Alibaba Cloud account or RAM user that triggered the change.
   Change Status	Possible statuses: In Progress, Complete, Interrupt Change, and Change is blocked. If the status is Change is blocked, click View Reason to see what is blocking the process.
   Actions	Contains the action buttons described below.
   View Change Records	Click View Change Records to open the Change Records page and view past change records. For more information, see View change progress and change records.
   Refresh	Click Refresh to get the real-time status of the instance change.

   View change stage details

   Click the icon next to a change task to view the progress of each of its stages:

   • Each stage shows the current progress and elapsed time. Completed stages also show their completion time.

   • In the Updating Nodes stage, the system displays the substages for each node's rolling update, including: Waiting for Cluster Status, Changing Node Specifications, Preparing Node Configuration, Starting service, and Waiting for Process Health.

     In the Task List panel, you can switch between the Change Progress and Data Migration Progress tabs. The change progress table displays Change Start Time, Change Type, Change Role, and Change Status information. Click Expand Details to view the task progress flow. For example, a create instance task includes three stages in sequence: Task Initialization, Scale Nodes, and Task Completion. Each stage displays a green checkmark and completion time after it finishes. Click View Logs to view the change logs.

   Important

   If the cluster status is unstable or unhealthy, the process may stall for a long time at the Waiting for Cluster Status or Waiting for Process Health stage. To ensure the task proceeds smoothly, reduce read and write operations from your application.

   Actions column

   • Show Details: View the specific configuration information for this change.

   • View Logs: Go to the View Logs page to view the instance's operation logs. For more information, see Query logs.

   • Pause Change / Resume Change (applies only to instances with changes in progress): Click Pause Change to pause the change task. After pausing the task, click Resume Change to continue it.

     Important

     • If an instance is in the Interrupt Change state, cluster services may be affected. You can resume the task by clicking Resume or by initiating a supported subsequent change, such as a cluster scale-up or plug-in management task.

     • After you click Resume Change, the entire restart process runs again, restarting the cluster nodes.

     • Only instances deployed in the Basic Control Architecture (v2) deployment mode support pausing changes. Instances deployed in the Cloud-native Control Architecture (v3) deployment mode do not. You can view the deployment mode in the Control Architecture Type section on the Basic Information page of the instance.

   • Cancel Change: Cancels an in-progress plug-in installation or uninstallation.

   Data migration progress

   Click the Data Migration Progress tab to view the progress details of the instance's migration task.

   Note

   • If an active data migration task is running for the instance, a red dot appears on the Data Migration Progress tab.

   • Migration progress details are displayed only when a cluster configuration change triggers a rolling restart that includes a data migration task.

   • The data migration progress shown in the console is not real-time. To view the precise migration progress, run the GET _cat/recovery command.

   The migration task progress details table contains the following columns:

   • Index Name

   • Source Node IP

   • Target Node IP

   • Status

   • File Count

   • File Recovery Progress

   • Data Size

   • Data Recovery Progress

   • Translog Operations

   • Translog Operation Recovery Progress

   If a migration task is active, you can click Adjust Migration Progress on the Data Migration Progress tab to adjust the following data migration parameters.

   Parameter	Elasticsearch parameter	Description
   For scenarios with a large number of shards
   Primary Shards Restored in Parallel	cluster.routing.allocation.node_initial_primaries_recoveries	The number of primary shards that can be recovered concurrently when a node restarts. If the node has multiple disks and the I/O pressure is low, you can increase this value.
   Data Balancing Tasks Run in Parallel	cluster.routing.allocation.cluster_concurrent_rebalance	The number of data rebalance tasks that can run concurrently within the cluster. This is suitable for scenarios where nodes are added or removed and the cluster load is not high.
   Data Restoration Tasks Run in Parallel	cluster.routing.allocation.node_concurrent_recoveries	The number of data recovery tasks that can run concurrently on a node, excluding primary shard recovery during a restart.
   Shards Restored to Node in Parallel	cluster.routing.allocation.node_concurrent_incoming_recoveries	The maximum number of concurrent shard recoveries allowed on a node. This typically refers to shards being restored to the current node from other nodes.
   For scenarios with a large amount of data
   Data Transmission Bandwidth	indices.recovery.max_bytes_per_sec	The maximum data transfer bandwidth for recovery on each node, for example, 100mb.

   Important

   • These adjustments are temporary and apply only during the change. Afterwards, the system restores the default values based on the cluster status.

   • The parameters listed above are dynamic settings for an Elasticsearch cluster. We recommend adjusting them to speed up data migration only when you have large amounts of data or slow recovery times. Adjust these parameters with caution. Frequent adjustments or settings that exceed the cluster's scheduling capacity can overload the cluster nodes.