Skip to main content

Create and manage parameter groups

Parameter groups allow you to view and manage parameters required for MemStore Redis OSS configuration based on engine version.
You can also manage parameters in bulk for all clusters linked to a specific parameter group.

The parameter list and core concepts in MemStore are compatible with Redis OSS. Since changing parameters may affect your MemStore service, it is highly recommended to review the appropriate values in advance.

Here is how to manage parameter groups in the MemStore service.

info

View parameter group list

You can view both default and custom parameter groups provided by KakaoCloud MemStore.

  1. Go to the KakaoCloud Console > Data Store > MemStore menu.

  2. Click the Parameter Groups menu, then choose the Custom Group or Default Group tab.

  3. Proceed to the matching Parameter Group menu.

    ItemDescription
    NameName of the parameter group
    Cluster modeWhether cluster mode is used
    FamilyMemStore engine version compatible with the group
    Linked clustersNumber of clusters linked to this group
    - A warning icon appears if any cluster has parameter sync issues
    DescriptionDescription of the parameter group
    MoreCustom groups only
    - Edit: Modify parameters
    - Copy: Create a new group with the same parameters
    - Reset to default: Revert parameters to default values
    - Rollback: Revert to previous parameter values (unavailable if no history or already rolled back)
    - Delete: Delete the parameter group

View parameter group details

You can check detailed information of a MemStore parameter group.

  1. Go to the KakaoCloud Console > Data Store > MemStore menu.

  2. Click the Parameter Groups menu, then choose the Custom Group or Default Group tab.

  3. Navigate to the appropriate tab and select the parameter group to view.

    ItemDescription
    NameName of the parameter
    ValueParameter value
    TypeParameter type
    - Some Dynamic parameters apply globally and require session refresh
    Data formatFormat of the parameter value
    EditableIndicates if the parameter is editable: No, Yes, or Yes (modified)

Create parameter group

  1. Go to the KakaoCloud Console > Data Store > MemStore menu.

  2. Click the Parameter Groups menu and select the Custom Group tab.

  3. In the Custom Group tab, click the icon and select Create Parameter Group.

  4. On the Create Parameter Group page, fill in the required details. Parameter values can be edited after creation.

    SectionFieldDescription
    Basic SettingsParameter group nameName of the parameter group
    - Must be unique within the same project
    Description (optional)Up to 100 characters
    Group SettingsFamilyMemStore family engine version
    Cluster modeWhether cluster mode is enabled
    Default parameter groupSelect from predefined groups for the engine version
    Parameters-Initial values for the parameters in the group

Modify parameter group

For custom parameter groups, you can edit the description and modify parameter values after creation. Only parameters with Editable: Yes can be modified. Once edited, the status changes to Yes (modified).

info
  • Some Dynamic parameters apply globally and require a session refresh.
  • If the sync status of a linked cluster is Pending or Applying, parameters cannot be modified.
  • Selected checkboxes persist across page navigation when editing.
caution
  • Some parameter modifications may affect MemStore service operation. See Parameters to be careful when modifying.
  • If issues occur after modification, consider using Copy, Retry, Rollback, or Reset to default options to manage parameter backups.
  1. Go to the KakaoCloud Console > Data Store > MemStore menu.

  2. Click the Parameter Groups menu and select the Custom Group tab.

  3. After identifying the group to modify, click the [More] icon > Edit.

  4. On the Modify Parameter Group page, update the desired fields.

    SectionFieldDescription
    Basic InfoParameter group nameName of the group
    Description (optional)Up to 100 characters
    Editable Parameters-List of editable parameters in the group
    - Modified parameters are marked as Yes (modified)
  5. If the parameter group is linked to multiple clusters, you can choose how to apply the changes:

    • Apply all at once: Apply the changes to all clusters simultaneously.
    • Apply sequentially: Apply the changes to each cluster one at a time.

Parameters to be careful when modifying

When modifying parameter values in a custom parameter group, some parameters may impact the operation of the MemStore service. Please refer to the following details.

info

repl-backlog

  • A buffer that temporarily stores data for partial synchronization (psync) when replication between primary and replica nodes is interrupted due to network issues.
  • During a network partition, the primary node stores data in the backlog and attempts psync once the replica reconnects.
  • The primary node retains backlog data for a certain time (repl-backlog-ttl) after the last replica disconnects. After that, the buffer is released.
  • If the replica reconnects after the buffer is cleared, a full sync is triggered.
  • Related parameters: repl-backlog-size, repl-backlog-ttl

client-output-buffer-limit-normal-hard-limit

TypeRangeDescription
integer0~Maximum client output buffer memory.
When set to 0, there is no limit.

Potential issue
If memory usage exceeds this limit, the client connection is closed.
Setting this value too low may prevent MemStore provisioning.
Recommended to use a sufficient value or default (0) based on workload.

client-output-buffer-limit-normal-soft-limit

TypeRangeDescription
integer0~Threshold for client output buffer memory.
Applies during a period defined by client-output-buffer-limit-normal-soft-seconds.

Potential issue
If memory usage exceeds the threshold during the defined time, the client connection is closed.
Setting this value too low may prevent MemStore provisioning.
Recommended to use a sufficient value or default (0).

client-output-buffer-limit-normal-soft-seconds

TypeRangeDescription
integer0~2147483647Allowed time (in seconds) for exceeding soft limit before disconnect.

Potential issue
If memory usage exceeds the soft limit for longer than this period, the client connection is terminated.
Recommended to use a sufficient value or default (0).

min-replicas-max-lag

TypeRangeDescription
integer0~2147483647Max ping delay allowed from replica to primary node.

Potential issue
If the primary doesn't receive a ping within this time, the replica is marked as unavailable.
Recommended to set an appropriate value based on service reliability needs.

min-replicas-to-write

TypeRangeDescription
integer0~2147483647Minimum number of replicas required for the primary to process write requests.

Potential issue
If the number of healthy replicas is less than this value, writes will be rejected.
Example: min-replicas-to-write: 5 with only two replicas results in:
primary.memstore:6379> SET hello memstore(error) NOREPLICAS Not enough good replicas to write.

repl-backlog-size

TypeRangeDescription
integer16384~Size of the backlog buffer for replication between primary and replicas.

Potential issue
If the replication lag exceeds the backlog size, a full sync will be triggered.
Setting this value too low may impact performance or cause temporary service unavailability.
Recommended to configure a sufficient value based on workload.

repl-backlog-ttl

TypeRangeDescription
integer0~2147483647Retention period (seconds) for backlog buffer in the primary node.
0 means never released.

Potential issue
After this time, the backlog is cleared and full sync is required.
Low values may cause frequent full syncs, affecting performance or availability.
Set a sufficient value based on workload.

maxclients

TypeRangeDescription
integer100~65000Maximum number of client connections per node.

Potential issue
Setting this too low may cause connection failures and trigger failover due to health check failures.
Configure a sufficient value based on node type and workload.

Copy parameter group

You can use the copy function to back up a custom parameter group by creating a new group with the same settings.

  1. Go to the KakaoCloud Console > Data Store > MemStore menu.
  2. Click the Parameter Groups menu and select the Custom Group tab.
  3. Find the group to copy, then click the [More] icon > Copy.
  4. In the Copy Parameter Group popup, review and confirm the information.
  5. Confirm that the copied group appears under the Custom Group tab.

Reset parameter group to default

You can reset a custom parameter group back to the default values based on the selected engine version. Resetting applies the default values to all linked clusters.

info

You cannot reset a group if any linked cluster is in Pending or Applying state.

  1. Go to the KakaoCloud Console > Data Store > MemStore menu.
  2. Click the Parameter Groups menu and select the Custom Group tab.
  3. Find the group to reset, then click the [More] icon > Reset to default.
  4. In the Reset to Default popup, review the notice and click Confirm.
    If linked to multiple clusters, choose how to apply:
    - Apply all at once: Applies changes to all clusters simultaneously.
    - Apply sequentially: Applies changes one by one. Clusters waiting are marked as Pending.
  5. Verify that the parameter sync status is updated correctly across all clusters.

Roll back parameter group

You can roll back a custom parameter group to the previous values only if there is a prior modification history.

info
  • Cannot roll back if any linked cluster is in Pending or Applying state.
  • Once rolled back, further rollback is not possible unless new changes are made.
  1. Go to the KakaoCloud Console > Data Store > MemStore menu.
  2. Click the Parameter Groups menu and select the Custom Group tab.
  3. Find the group to roll back, then click the [More] icon > Rollback.
  4. In the Rollback Parameter Group popup, review the notice and click Confirm.
  5. Verify that rollback is reflected correctly across all linked clusters.

Delete parameter group

You can delete a custom parameter group that is no longer in use.

info
  • Default parameter groups cannot be deleted.
  • Parameter groups linked to clusters cannot be deleted.
  1. Go to the KakaoCloud Console > Data Store > MemStore menu.
  2. Click the Parameter Groups menu and select the Custom Group tab.
  3. Find the group to delete, then click the [More] icon > Delete.
  4. In the Delete Parameter Group popup, enter the required information and click Confirm.
  5. Confirm that the group is deleted from the Custom Group tab.

Retry parameter application

You can retry applying parameters only when the parameter sync status of the linked instance is not In-Sync, but Error-Sync or Suspended.
For more details about parameter sync statuses, refer to Parameter application status.

  1. Go to the KakaoCloud Console > Data Store > MemStore menu.
  2. Click the Parameter Groups menu and select the Custom Group tab.
  3. Navigate to the appropriate Parameter Group menu and select the parameter group to view its details.
  4. Go to the Linked Clusters tab.
  5. Click the [Retry] button for the instance where you want to retry the application.