OpenText Analytics Database 26.2.x – Database management with VCluster UI

Admin: Dashboard

Mon, 01 Jan 0001 00:00:00 +0000

The VCluster UI Dashboard is the default landing page of the interface. It provides quick access to health metrics and other key sections allowing you to navigate and manage your database efficiently.

Access Health Metrics

Click Health Metrics. The Connect Vcluster to Prometheus modal appears.
Provide OpenText™ Analytics Database's embedded HTTPS server port number and click Start Prometheus. The default is 8443.
Click Accept to authorize the Prometheus connection.
Click Confirm. VCluster is now connected to Prometheus to scrape metrics.

You will be directed to the Dashboard which displays health metrics such as CPU Usage by Node, Memory Usage by Node, Disk I/O Usage by Node, and Free Disk Space by Node. Click any chart to open a modal where you can filter by time and node.
You can set the Refresh Interval to automatically update the health charts at your preferred frequency.
Click Other Health Data to access Health Watchdog and Cluster Health pages.
Clicking Health Metrics keeps the dashboard in health metrics view during your browser session. To go back to the default dashboard, click Stop Health Metrics and then click Confirm.

Important

Health metrics only work if you have enabled them by navigating to port 9090 and accepting the certificate prompt.

Admin: Subcluster

Mon, 01 Jan 0001 00:00:00 +0000

The VCluster UI Subcluster page displays a list of your subclusters and the number of nodes in each subcluster. You can also add, remove, start, stop, sandbox, and unsandbox subclusters.

Node Details for a subcluster enables you to view the node name, address, state, catalog path, and primary node.

Admin: Database

Mon, 01 Jan 0001 00:00:00 +0000

The Database page shows you the name of your database along with the nodes, IP Address, and status. You can also view more details for each node such as node name, catalog path, data path, depot path, and communal storage location.

The following options are available on this page:

Note

If you do not have an existing database, only the Create Database and Revive Database options are available.

Start Database
Stop Database
Re-IP Database
Drop Database
Revive Database: You will not see this option if a database already exists.
Additional Options: Include Upgrade License and Manage Packages.
- The Upgrade License option enables you to choose your license file and upgrade to it.
- The Manage Packages option enables you to install or uninstall pacakges. It lists all the packages that are installed or uninstalled. You will be directed to the Job Status page when you install or uninstall packages.

Admin: Nodes

Mon, 01 Jan 0001 00:00:00 +0000

The Nodes page shows you all of your database nodes. You can add, remove, start, and stop nodes on this page.

Note

The Add Node screen lets you choose between Regular and Compute node types. This option is available only for databases running in Eon Mode.
You can select any node and click Stop Node. In the Stop Nodes screen, you can enable Force Kill which terminates the node using SIGKILL instead of a graceful shutdown. Use this option only if a graceful shutdown fails as it may result in data loss.

Admin: Job Status

Mon, 01 Jan 0001 00:00:00 +0000

The Job Status page shows you the status of your database jobs and the Workload Replay captures and replays:

Database

Database shows you the status of jobs from the Database, SubCluster, and Nodes pages. This includes adding a node or subcluster, changing the IP of the database, creating the database, and the other options.

Workload Replay

Admin: Workload Replay

Mon, 01 Jan 0001 00:00:00 +0000

The Workload Replay page lets you view your database workload by viewing database queries and how long they take.

This page is used for the following:

Evaluating the performance of a new database version before deploying it to the production environment.
Assessing changes to the system configuration or test the database with new knobs or code changes.
Assessing the performance of upgrading hardware components or switching to a different operating system.

It uses three tabs to help you perform the assessment: Capture Workload, Replay Workload, and Reports.

Capture

Capture is used to capture a query or multiple queries to see their performance. You specify a start and stop timestamp when the query/queries you want to capture were run. Then click Capture Worload to capture the query/queries. You are redirected to the Job Status > Workload Replay page to see the status of the capture. The job name will be capture-workload and once the status shows successful you can replay the workload. You need to note the file path for the capture job to run the replay.

Replay Workload

After running a capture, Replay Workload is used to compare the first query run time with a new query run time. Select a sandbox from the dropdown to use for the replay. If no sandbox is present in the dropdown, you can create a sandbox using Create sandbox. You need to know the file path from the Job Status > Workload Replay page to select the capture you want to run against. Select that in the Workload File drop down. In the Sandbox drop down, select the sandbox you are using for the replay, then click Replay Workload. You are redirected to the Job Status > Workload Replay page to see the status of the capture.

The job name will be replay-workload and once the status shows successful you can replay the workload. You need to note the file path for the replay job to run the report.

If you want to cancel the replay that is running, hit the Cancel Replay Workload button.

Reports

After running the replay, Reports lets you view the query/queries from both the capture and the replay side by side. In the Workload Replay Report File, make sure the replay file you want to view is selected. It can be found on the Job Status > Workload Replay page. You can see a chart of the original time and the replay time which can be customized to how you want to view it. Scrolling down also shows you a table of the results.

Admin: Cluster Health

Mon, 01 Jan 0001 00:00:00 +0000

The Cluster Health page lets you monitor the health of your database. There are three sections on this page:

Event Detailed Info - displays the details of an event if you click on it in the Slow Event Duration Over Time chart. You can click on the event details to see additional details about the event.
Events Statistics - displays the number of slow events, the longest and average event time, and lets you run a manual slow event check over a specified time period.
Slow Event Duration Over Time - displays the events from the preview day, 7 days, or 30 days (per your selection) in a chart.

Admin: Health Watchdog

Mon, 01 Jan 0001 00:00:00 +0000

The Health Watchdog page lets you monitor your database using Health Watchdog. It displays the health of your database and mergeout queue size, lets you set configuration parameters for your database, and view blocked transactions.

Database Status

The Database Status page shows you your cluster state and has charts to track the different health metrics. You can set the chart to display recently elapsed time window or for a fixed time range.

Configuration Parameters

The Configuration Parameters page lets you configure the health metrics for your database. It displays the current value for the health metric and lets you input new values if you need to adjust your settings. You can also select whether transactions are blocked or not.

Blocked Transactions

The Blocked Transactions page lets you view any transactions that have been blocked and lets you cancel any transactions that you select.

Admin: Certificate Management

Mon, 01 Jan 0001 00:00:00 +0000

The Certificate Management page shows you your certificate information and manage your certificates.

The following options are available to manage certificates:

Pending Certs list: Shows a list of the status of server and client certificates.
Issued Certs list: Shows a list of the issued certficates.
Revoked Certs list: Shows a list of the revoked certificates.
Renew Server Cert: Lets you renew your server certificate.
Renew Client Cert: Lets you renew your client certificate.
Apply Client Cert: Use this option to apply for a client certificate.

Admin: Query Profile

Mon, 01 Jan 0001 00:00:00 +0000

The Query Profile page allows you to import a tar file containing query profile data into tables. After importing, the Job Status page displays the target schema and key ID. You can then use this schema and key ID in the LLM to analyze your queries.

Import query profile data

Navigate to the Query Profile page.
Click Browse to select the tar file you want to import.
Enter a Target Schema name where the query profile data will be imported.
(Optional) Enter a Key ID to identify the profile data.

Note
A Key ID is a unique string that helps you keep record of your stored profiles. If you do not provide a key ID, a unique ID will be automatically generated. In the backend, copies of the stored internal tables are created using the naming format: <target_schema>.qprof_<table_name>_<key_id>. The combination of target schema and key ID can be used by the LLM to locate your saved profiles.
(Optional) Enter a Sandbox Name for the import job.

Note
If you do not provide a sandbox name, the profiles will be imported to the main cluster.
Click Import to begin uploading the data. The import job is created and you are directed to the Job Status page.

Monitor import job status

On the Job Status page, in the Query Profile tab, locate your import job in the list.
Once the job completes, note the Target Schema, Key ID, and Sandbox Name (if provided) displayed in the job details.
Use the target schema, key ID, and sandbox name (if applicable) with the MCP Server Query Profiling tools to leverage the LLM to analyze your imported query profiles.