Feature Summary and Revision History

Feature Description

A new REST API is introduced to query shard information of given session or binding.

• API details: https://{DRA_MASTER_IP}/dra/api/binding/shardDetails/{dbName}/{searchKey:.+}

  where,

  DRA_MASTER_IP - DRA VNF IP

  dbName - Any one of the following values:

  • session

  • ipv4

  • ipv6

  • imsi

  • msisdn

  searchKey:.+ - Session or binding value

The following new statistics is added:

• dra_api_binding_sharddetails_count

For more information on statistics, see Statistics/KPI Additions or Changes.

Feature Summary and Revision History

Feature Description

CPS now supports configuring realm in Policy Builder relay endpoints.

A new column with name Realm is added to support configurable relay endpoint realm name.

For more information, see Policy DRA Relay Configuration section in the CPS vDRA Configuration Guide.

Feature Summary and Revision History

Feature Description

vDRA supports rebalancing the connections across the DRA Directors to make sure that if a signaling spike occurs there is CPU available to process the messages.

In order to manually rebalance DRA, you must know which connections are on which DRA Directors. This helps to figure out which connections should be disconnected to allow them to reconnect on the DRA Director with the lower number of connections.

Currently, you need to configure the Director ID manually.

Here is a sample configuration:

To support extended peer monitoring, existing REST APIs are enhanced to return the peers connected to each Director instance.

• Existing “activePeerEndpoints” REST API is enhanced with additional field “instanceId”.

  • API details: https://{DRA_MASTER_IP}/dra/api/activePeerEndpoints

• Existing “localActivePeerEndpoints/disconnect/key/{searchKey:.+}” REST API is used to gracefully disconnect a peer connection.

  • API details: https://{DRA_MASTER_IP}/dra/api/localActivePeerEndpoints/disconnect/key/{searchKey:.+}

    Value to be provided in {searchKey:.+} is the key value which is returned in API call “activePeerEndpoints”.

Feature Summary and Revision History

Feature Description

vDRA now supports collecting all the mongod log files from different VMs and create a single consolidated MongoDB log file based on the start and end timestamps which can be provided as inputs.

This feature also provides flexibility to get the consolidated view of all the MongoDB logs into a single file or collect as a singletar.gz file for offline analysis.

For more information, see the following sections in the CPS vDRA Operations Guide.

• debug collect-db-logs-advanced collect

• debug collect-db-logs-advanced scan

Limitations

• This utility can fetch logs only for the mongod instances which are running on respective sites where commands are executed.

• The log collection is limited to 15 days. If you need logs beyond 15, you must login to VM directly to pull the logs.

• Before executing debug collect-db-logs-advanced scan command, you need to execute collect command which pulls all the logs from different VMs into tar.gz.

• debug collect-db-logs-advanced scan command allows you to input timestamps in maximum of 6 hours time interval. Currently, this command expects tar.gz file to be present in the respective storage location and creates consolidated-log-output in same place.

Feature Summary and Revision History

Feature Description

vDRA now has the ability to throttle incoming request based on database VM CPU usage. If database VM CPU crosses threshold value, calls are rejected.

You can define CPU threshold for read database operations for write DB operations. For CLI command, see binding throttle-db-operation section in the CPS vDRA Operations Guide.

If bindings are best effort then calls are processed without performing any DB operation if CPU usage threshold is breached.

To support the feature, in Error Result Code Profile CRD under Error field, new error code DB CPU Overload is added.

Error profile can be configured with this value and result code that needs to be sent in response to PCEF for messages which are rejected due to DB VM CPU overload.

The following new statistics has been added:

• db_cpu_threshold_breach_total.

For more information on statistics, see Statistics/KPI Additions or Changes.

Feature Summary and Revision History

Feature Description

DRA provides support to generate alerts when the docker container state is un-healthy. The alert is resolved immediately after the container state changes to healthy.

Feature Summary and Revision History

Feature Description

vDRA now uses memory-based volume (“tmpfs”) in health monitoring services (keepalived, keepalived-monitor, docker-host-info, etc.) to store/track the health status information of the monitored services.

It also support migration to “tmpfs” volume upon software upgrade and fresh installation. Downgrade reverts the system back to using disk-based volumes.

Memory and Performance Impact

As tmpfs volumes are created in memory, there is an increase in the memory usage by monitoring services.

Upgrade/Migration/Backward Compatibility Considerations

Note	The upgrade from non-tmpfs version to tmpfs version would disrupt the service. It is recommended that the traffic is directed to an alternate site when performing upgrade. For more information, contact you Cisco Account representative.

The feature is enabled when the system is upgraded through normal upgrade procedure.

When upgrading DRA VNF, it migrates the required health monitoring services to use tmpfs volume for health status configuration and status information.

vDRA supports migrating the required health monitoring services to use tmpfs volume for health status configuration and status information. Downgrade reverts the system back to using disk-based volumes.

Feature Summary and Revision History

Feature Description

vDRA now uses Azul C2 compiler JVM for director and worker nodes.

Feature Summary and Revision History

Feature Description

CPS now provides support to generate alarm based on Docker Engine state when VM goes down.

The following new alarm is added:

• DOCKER_ENGINE_DOWN

For more information, see Component Notifications and Sample Alert Rule Configuration sections in the CPS vDRA SNMP and Alarms Guide.

Feature Summary and Revision History

Feature Description

vDRA now uses latest stable docker engine version 19.03. The latest version includes fixes for open issues, security patches and some improvements.

You can check the docker by using the docker --version command from VM shell.

For more information, see https://docs.docker.com/engine/release-notes/.