REDREDGROUP Document Hub
What’s New
RDC CloudMemory Database

Manage and delete a Redis database

Control runtime state and remove an unused memory database.

Overview layout

The summary at the top shows runtime status, environment, region, and instance type. Available charts use two columns on larger screens and one column on small screens, with time and value axes. The Details table below starts with Instance ID and Resource ID, with copy buttons for these identifiers.

View resource metrics

Open Overview and review Metrics. You need project VIEW access or higher, or workspace owner access.

ChartMeaning
Memory usageRedis allocated memory (used_memory), its configured maxmemory limit, and the latest usage percentage. This is not the container's total resident memory.
ConnectionsConnected clients and Redis's maxclients limit. Monitoring connections are excluded.
CPU usageRedis process CPU time consumed between consecutive samples, divided by elapsed time and the plan's allocated vCPU. This excludes child processes and is not host-wide or total-container CPU usage.

Select Last hour, Last 6 hours, Last 24 hours, or Last 7 days. Hover over a line to inspect its time and value. Faint lines show limits. The values above each chart show the latest individual sample in the selected period.

Metrics are collected approximately once a minute from active resources and are available for seven days. Charts use one-minute intervals for one and six hours, five-minute averages for 24 hours, and 30-minute averages for seven days. Averages include successful samples only; brief peaks can be less visible in longer ranges.

Select Refresh to retrieve stored samples. The page does not poll automatically, and refreshing does not start collection. Hover over a chart point to see its timestamp and value. A warning indicates when the displayed measurements are out of date.

Missing intervals are left empty, not filled with zero. CPU requires two consecutive samples and is unavailable on the first sample after a restart, a collection gap, or a CPU allocation change. A zero maxmemory setting means Redis has no configured memory limit, so its usage percentage and limit line are unavailable.

New databases have no historical samples. If graphs remain empty while the resource is active, contact support with the resource ID and affected time range. Threshold rules and email alerts are not part of these charts yet.

Memory allocation and limits

To reduce the risk of the database stopping because it runs out of memory, RDC Cloud sets Redis's usable memory limit (maxmemory) to 75% of the instance memory included in your plan. The remaining 25% provides operating headroom for allocator overhead, client buffers, and background work.

For example, a 512 MiB instance has a 384 MiB Redis limit and 128 MiB of operating headroom. The Redis limit includes memory overhead, so it is not a guarantee that you can store 384 MiB of application data. The reserve reduces the risk of reaching the container's memory limit but cannot prevent every out-of-memory event.

The noeviction policy rejects writes that need additional memory once Redis reaches its limit, instead of evicting existing keys. Applications should handle these errors and expire or remove unneeded keys.

The memory limit persists across restarts. For an existing database without this limit, use Console Start or Restart to apply it; a restart interrupts active connections. The chart shows the actual limit after the next successful sample. Older samples retain the limit that applied when they were collected.

Start, stop, or restart

Use Start, Stop, or Restart on the resource detail page. Wait for an active operation to complete before submitting another. Check Overview for the resulting runtime state and test a client connection afterward.

Stopping makes the instance unavailable to clients; restarting can interrupt active connections. Neither is a substitute for exporting data before deletion.

Delete the database

Deletion removes the database and its stored data. Export anything you need and update dependent applications first.

  1. Open Danger Zone.
  2. Select Delete.
  3. Review the resource name and confirm with Delete memory database.
  4. Verify that the resource is no longer active in the project.

Deletion cannot be undone. Use Stop when you only need to pause the workload.

Network traffic

View Network Inbound and Network Outbound in the resource details. See Network traffic for collection intervals, permissions, and how to interpret gaps.

Passkey confirmation for deletion

If your account has a registered passkey, deleting this cloud resource requires a new passkey verification, even if you already verified your sign-in. After confirming deletion in Console, use Verify with passkey in the RDC Account window. Allow pop-ups for Console if the verification window is blocked.

The verification applies only to the selected resource and one deletion attempt. Cancelling, closing the window, or letting the request expire does not send the deletion request. If deletion fails, start a new attempt and verify again. Accounts without a registered passkey continue to use the existing deletion confirmation. Your resource deletion permission is still required.

On this page