Table of Contents

General Feature Release 10.5.5 – Preview

Important

We are still working on this release. Some release notes may still be modified or moved to a later release. Check back soon for updates!

Important

Before you upgrade to this DataMiner version, make sure version 14.40.33816 or higher of the Microsoft Visual C++ x86/x64 redistributables is installed. Otherwise, the upgrade will trigger an automatic reboot of the DMA in order to complete the installation.

The latest version of the redistributables can be downloaded from the Microsoft website:

Tip

Highlights

No highlights have been selected yet.

New features

SNMP forwarding: New option to prevent an SNMP manager from resending SNMP Inform messages [ID 41884]

Up to now, when you stopped and restarted an SNMP manager, all open alarms would be resent. From now on, when you configure an SNMP manager, you will be able to prevent this by selecting the Enable tracking to avoid duplicate Inform Acknowledgments (ACKs) option. If you select this option, DataMiner will track which Inform messages have been sent, and will not resend those that have already been acknowledged.

Note

This new Enable tracking to avoid duplicate Inform Acknowledgments (ACKs) option is not selected by default and is not compatible with the existing Resend all active alarms every: option. It is also not compatible with the Resend... command, which in DataMiner Cube can be selected after right-clicking an SNMP manager in the SNMP forwarding section of System Center.

Changes

Enhancements

DataMiner recycle bin enhancements [ID 40565]

The C:\Skyline DataMiner\Recycle Bin\ folder contains backup copies of modified configuration files and folders, stored as zip files. Each zip file includes the modified file or folder along with a Cause.txt file, which details the reason for the change and its timestamp. These backup copies help you restore previous configurations if needed.

Up to now, a separate zip file would be created for each configuration change that had been implemented in the system.

From now on, the contents of the C:\Skyline DataMiner\System Cache\Recyclable\ folder will be zipped and moved to the C:\Skyline DataMiner\Recycle Bin\ folder every 11 minutes. This process will first occur 3 minutes after DataMiner startup.

When a configuration change occurs, two scenarios are possible:

  • If the file or folder has not been modified after the most recent move to the Recycle Bin folder (which happens every 11 minutes), a new entry is created in the C:\Skyline DataMiner\System Cache\Recyclable\ folder with the name of the changed file or folder.

  • If the file or folder has been modified after the most recent move to the Recycle Bin folder, the existing entry in the C:\Skyline DataMiner\System Cache\Recyclable\ folder is not replaced. Instead, the Cause.txt file is updated with the new change description and corresponding timestamp.

RecycleBinSize setting in MaintenanceSettings.xml

In the MaintenanceSettings.xml file, you can specify the maximum size (in MB) of the DataMiner recycle bin in the RecycleBinSize tag.

From now on, the system will check every 7 minutes whether storage limits have been exceeded. If the system detects a breach, it will perform a cleanup on the recycle bin to restore the storage within acceptable limits:

  • If the number of files exceeds the limit, the system will clean up the recycle bin until it holds 80% of the lowest value between the maximum allowed number of files (default: 1000) and the current number of files.

  • If the folder size exceeds the limit, the system will clean up until the folder size is no longer over the configured size limit.

This cleanup will occur for the first time 2 minutes after DataMiner startup. Up to now, the recycle bin was cleaned to the maximum size and number of files every hour.

Note
  • Whatever the maximum size specified in the RecycleBinSize tag, the maximum number of files in the recycle bin is limited to 5000.
  • The default recycle bin size is 100 MB.
  • From now on, if the recycle bin size is set to 0 MB or an invalid size, it will revert to the default value of 100 MB.
Restoring a previous configuration

If an incorrect configuration change is implemented in the system, in some cases, it is possible to use the recycle bin to restore the original configuration.

For example, if a view is renamed or moved in the Surveyor, a zip file will be created containing the Views.xml file and a cause file, which details why the change occurred. It is then possible to restore the Views.xml file as follows:

  1. Copy the file from the Recycle Bin folder back to its original location.

    Note

    From now on, the files in the Recycle Bin folder are only updated every 11 minutes. This means that when you restore the files, they may not contain recent changes.

  2. Restart the DMA.

  3. Force a synchronization of the file in the DMS.

Caution

Always be extremely careful when using this functionality, as it can have far-reaching consequences on your DataMiner System.

BPA test 'Check Deprecated MySQL DLL' renamed to 'Check Deprecated DLL Usage' [ID 42057]

Up to now, the BAP test named Check Deprecated MySQL DLL would check whether the MySql.Data.dll was not outdated.

Now, this BPA test has been renamed to Check Deprecated DLL Usage. Depending on the DataMiner version, it will checks for the following DLL files, in the specified folders:

Deprecated DLL Deprecated since DataMiner version Minimum safe DLL version Folder
MySql.Data.dll 10.4.6/10.5.0 8.0.0.0 C:\Skyline DataMiner\ProtocolScripts
SLDatabase.dll 10.5.5/10.6.0 N/A C:\Skyline DataMiner\ProtocolScripts or C:\Skyline DataMiner\Files

Any version lower than the specified minimum version will be considered outdated, as older versions are known to pose security risks.

Service & Resource Management: UpdateReservationInstanceEventHasRun method will now first clone the reservation object before updating it [ID 42124]

Up to now, the UpdateReservationInstanceEventHasRun method would directly update the cached reservation object and then save it.

From now on, it will clone the reservation object in the cache, make the change, and then update the cached object.

Security enhancements [ID 42307]

A number of security enhancements have been made.

DataMiner Object Model: An error will now be returned when a FieldValue was added for a non-existing FieldDescriptor [ID 42358]

When, while a DOM instance was created or updated, a FieldValue was added for a non-existing FieldDescriptor, up to now, no error would be returned.

From now on, the trace data will indicate that a DomInstanceError was thrown with error reason FieldValueUsedInDomInstanceLinksToNonExistingFieldDescriptor.

Reduced memory usage when updating a large number of parameters in bulk [ID 42385]

When a large number of parameters are updated in bulk, from now on, SLProtocol will send the parameter changes to SLElement in chunks of 1000 rows. This will considerably reduce overall memory usage during serialization, especially when a large number of rows are updated due to e.g. aggregation or merge actions.

STaaS: An alarm will now be generated when a data type is being throttled [ID 42387]

If your system is pushing too much load for a specific data type, that data type will be throttled. This could for example happen when you have an element that is continuously saving parameter updates.

From now on, when this happens, an alarm will be generated with information about the data type or types that are being throttled.

GQI DxM will now shut down faster [ID 42428]

Because of a number of enhancements, the GQI DxM will now shut down faster, especially in situations where NATS is not running.

SLAnalytics - Relational anomaly detection: Input validation when adding a new parameter group [ID 42429]

When you add a new parameter group, from now on, an error message will appear when that parameter group contains

  • an invalid group name,
  • an invalid number of parameters,
  • an invalid anomaly threshold, or
  • an invalid minimum anomaly duration value.

Enhanced performance when restarting HTTP elements in a timeout state [ID 42443]

Because of a number of enhancements, overall performance has increased when restarting HTTP elements in a timeout state.

Protocols: HTTP status code will only be logged when debug logging level 4 is enabled [ID 42447]

Up to now, each time you opened Stream Viewer while HTTP communication was in progress, the response code would be logged to the element log.

From now on, the response code will only be logged to the element log if debug logging level 4 is enabled (regardless of whether Stream Viewer is open).

SLLogCollector now collects the output of the 'dotnet --list-runtimes' command [ID 42448]

SLLogCollector packages will now include the output of the dotnet --list-runtimes command.

The output will be stored in the following file:

\Logs\Windows\.NET runtimes\cmd.exe _c dotnet --list-runtimes.txt

GQI DxM will now look for missing dependencies in the Automation script libraries folder [ID 42468]

GQI extensions use the Automation engine to create DLL libraries that are then loaded by GQI to add ad hoc data sources, custom operators, etc.

GQI will now look for missing dependencies in the C:\Skyline DataMiner\Scripts\Libraries folder. This will allow GQI extension scripts to find the Automation script library at runtime.

Important

If the referenced Automation script library has dependencies of its own, these will also need to be added as dependencies in the GQI extension scripts.

GQI recording removed from GQI DxM [ID 42470]

GQI recording, a debugging feature that allowed you to save GQI communication and replay it in a lab environment, has now been removed from the GQI DxM.

SLAnalytics - Relational anomaly detection: New API messages [ID 42480]

The following messages have been added to the Relational Anomaly Detection API. These messages can be used to create, retrieve and remove RAD parameter groups:

Message Function
AddRADParameterGroupMessage Creates a new RAD parameter group.
If a group with the same name already exists, no new group will be added. Instead, the existing group will be updated.
GetRADDataMessage Retrieves the anomaly scores over a specified time range of historical data.
GetRADParameterGroupInfoMessage Retrieves all configuration information for a particular RAD parameter group.
GetRADParameterGroupsMessage Retrieves a list of all RAD parameter groups that have been configured.
RemoveRADParameterGroupMessage Deletes a RAD parameter group.
RetrainRADModelMessage Retrains the RAD model over a specified time range.
Note
  • Names of RAD parameter groups will be processed case-insensitive.
  • The following messages have been deprecated: AddMADParameterGroupMessage, GetMADParameterGroupInfoMessage, RemoveMADParameterGroupMessage, and RetrainMADModelMessage.

GQI DxM: Record limit of Sort operator has been increased to 100,000 records [ID 42492]

When the GQI DxM is being used, the record limit of the Sort operator will now be 100,000 instead of 10,000.

SLNetClientTest - DataMiner Object Model: Enabling debug logging [ID 42504]

In the DataMiner Object Model window, which allows you to see all details of a particular DOM module, a new Debug Logging section has now been added to the Maintenance tab. In this section, you can find the following buttons:

Button Description
Enable Adds or updates an override for the log file of the current DOM manager, setting all log levels to 6.
Reset Removes the override for the log file of the current DOM manager is removed, regardless of the tool that added it.

Also, a status label will now indicate whether debug logging is enabled or disabled.

Note
  • The above-mentioned status label will show "Enabled" when a level-6 override is present. If all log files have level 6 by default, the status label will show "Disabled" until you add an override.
  • Enabling debug logging may significantly increase the amount of logging that is written to disk.

GQI DxM: New life cycle method allows ad hoc data sources to optimize sort operators [ID 42528]

A new optional life cycle method has been introduced for ad hoc data sources running in the GQI DxM. It will allow to optimize or modify sort operators added to the query.

You can use this life cycle by implementing the Skyline.DataMiner.Analytics.GenericInterface.IGQIOptimizeableDataSource interface, which has one method:

IGQIQueryNode Optimize(IGQIDataSourceNode currentNode, IGQICoreOperator nextOperator)
  • currentNode is the query node that represents the current ad hoc data source.
  • nextOperator represents the next operator appended to the query.

This method should return the query node that represents the result of applying the next operator to the current ad hoc data source node. Similar to the custom operator implementation, the ad hoc data source implementation can decide to do the following:

  • Append the nextOperator to the currentNode (i.e. the default behavior when this life cycle method is not implemented).
  • Remove/ignore the nextOperator, usually taking responsibility of the operation internally.
  • Modify/add operators.
Note
  • This life cycle method will only be called when the nextOperator is a filter or a sort operator.
  • This life cycle method can be called multiple times if there is a new nextOperator.

New log viewer web page [ID 42533]

Up to now, client applications like DataMiner Cube used the ViewLog.asp web page to display server-side log files. This web page has now been replaced by the ViewLog.aspx web page.

This new log viewer page has improved compatibility with Failover setups and better error handling for HTTPS certificates.

Swarming: New 'Block Swarming' option to indicate that an element is not allowed to swarm [ID 42535]

When you create or update an element in DataMiner Cube, you will now be able to indicate that the element is not allowed to swarm to another host.

To do so, go to the Advanced section, and enable to Block Swarming option. By default, this option will be set to false.

If you try to swarm an element of which the Block Swarming option is set to true, then the error message Element is not allowed to swarm (blocked) will be displayed.

Note

In DataMiner Cube, this Block Swarming option will only be visible if Swarming is enabled in the DataMiner System.

DataMiner IDP license notice will no longer appear [ID 42574]

Since DataMiner version 10.1.0/1.0.0.8, the following notice would appear when a DataMiner Agent that was not using Indexing Engine had an IDP license but no ServiceManager license:

DataMiner IDP is licensed, but no Elasticsearch database is active on the system. Therefore, scheduled workflows are not available.

As DataMiner IDP no longer requires neither a separate license nor an Indexing Engine, from now on, this notice will no longer appear.

SLAnalytics: An anomaly alarm event will now be generated when a change point with a type that is not monitored is changed to a change point with a type that is monitored [ID 42596]

From now on, an anomaly alarm event will be generated when a change point with a change point type for which no anomaly monitoring was configured, is updated to a change point with a change point type for which anomaly monitoring is configured.

Fixes

Problem with aggregation alarms on Cassandra Cluster and STaaS [ID 42095]

Up to now, aggregation alarms would not work as intended on DataMiner Systems using a Cassandra Cluster database or Storage as a Service (STaaS).

DataMiner upgrade: No progress updates when DataMiner Cube was connected to the local DataMiner Agent [ID 42114]

When you either uploaded a DataMiner upgrade package or executed a DataMiner upgrade on a DataMiner Cube that was connected to the local DataMiner Agent via localhost, up to now, you would not get any progress updates regarding the upload or upgrade process.

Problem with SLElement when a DVE child element is deleted while the DVE main element is in a timeout state [ID 42137]

In some rare cases, an error could occur in SLElement when a DVE child element was deleted while the DVE main element was in a timeout state.

Web visual overviews: Incorrect load balancing when users were a member of the same groups [ID 42291]

When load balancing is implemented among DataMiner Agents in a DMS for visual overviews shown in web apps, DataMiner checks the group memberships of users to determine where to send requests to balance the load. However, when multiple connected users were a member of the same groups, updates for these users were not consistently handled by the same Agent, which could cause the load balancing to be less effective and more memory to be used.

SLDataGateway: Problem when some of the Cassandra nodes are marked as down [ID 42384]

In some cases, SLDataGateway could incorrectly get stuck in a state where some of the Cassandra nodes are marked as down.

Additional polling has now been introduced that will kick in when Cassandra nodes have been marked as down for a longer period. When those nodes prove to be up and running, they will forcefully be marked as up.

Element card of a DVE or Virtual Function could show incorrect alarm colors [ID 42402]

In some cases, the ParameterChangeEvent or ParameterTableUpdateEventMessage for a primary key cell would contain an invalid InstanceAlarmlevel or CellBubbleUpLevel. As a result, when you opened an element card of a DVE or a Virtual Function, the card pages would show incorrect alarm colors.

STaaS: Problem with logger table queries incorrectly not yielding any results [ID 42408]

When a logger table was queried, in some cases, the query would incorrectly not yield any results due to a filter conversion issue.

SLAnalytics: Problem when starting behavioral anomaly detection due to caching issue [ID 42422]

Up to now, in some cases, behavioral anomaly detection would not be able to start up correctly due to the maximum cache size having been reached when the internal caches were being fetched after SLAnalytics had been started.

From now on, if the maximum cache size is reached, old model information might get discarded to allow behavioral anomaly detection to start up correctly. If this happens, the following error will be logged:

Max cache size reached during prefetch of the cache, potential data loss

Problem when deleting an element or service property [ID 42434]

When an element property or a service property was deleted, in some cases, a KeyNotFoundException exception could be thrown, especially when one of the elements that had a value for that property was a migrated or a swarmed element.

The property would be deleted from the PropertyConfiguration.xml file, but not from the elements or services that had a value for the property in question.

'Unable to set the destination IP' errors would be logged in SLErrors.txt when setting up an SNMPv3 connection [ID 42453]

When an SNMPv3 connection was set up, log entries similar to the example below would be added to the SLErrors.txt file:

RT_QACTIONS_SNMP_v3: Unable to set the destination IP: polling IP=::1; resolved IP=::1; or=APPLY SECURITY FAILED: EMPTY USER NAME

GQI DxM: Ad hoc data source or custom operators would not receive SLNet events when using subscription sets [ID 42454]

When the GQI DxM was being used, ad hoc data sources and custom operators would not receive any SLNet events if they added or updated a subscription for a specific subscription set.

Problem when creating the ClusterEndpoints.json file [ID 42481]

In some cases, an error could be thrown when the ClusterEndpoints.json file was created in the C:\Skyline DataMiner\Configurations\ folder.

SLAnalytics - Relational anomaly detection: GetMADParameterGroupInfoMessage would incorrectly return an empty display instance for parameters with non-advanced naming [ID 42508]

For parameters with non-advanced naming, the GetMADParameterGroupInfoMessage would incorrectly return an empty display instance.

DVE settings could get out of sync with the element data when DataMiner or an element was restarted [ID 42515]

When DataMiner or an element was stopped while there were still DVE parameter sets in the queue, up to now, the DVE settings would be out of sync with the element data when DataMiner or the element was restarted. From now on, the DVE settings will be re-applied in the element data.

Note

Always make sure element names are unique, especially when using the noelementprefix option.

Incorrect 'Detected duplicate DVE' notice would appear due to a caching issue [ID 42546]

Due to a caching issue, in some cases, the following notice could incorrectly appear in the Alarm Console:

Detected duplicate DVE, did not create DVE for {DVE name}

Connection issue between SLSNMPManager and SLNet [ID 42547]

When multiple SLSNMPManager processes tried to simultaneously connect to SLNet, or when an element with multiple SNMP connections tried to connect to SLNet, in some cases, the following error would be logged in the Windows log Application:

(SLNetCOM SLSNMPManager.exe) Failed to connect to SLNet: (Code: 0x800402CD) Skyline.DataMiner.Net.Exceptions.DataMinerCommunicationException: Connection was closed at 15:37 (There's a new connection for this module/agent.)

GQI DxM would leak memory when an SLNet connection was disconnected [ID 42592]

When an SLNet connection created by the GQI DxM was disconnected, up to now, the associated compatibility manager would not be cleaned up correctly, causing the GQI DxM to leak memory.

Information events would incorrectly not get flushed to the database when an element was stopped [ID 42604]

When an element was stopped, contrary to alarm events, information events would incorrectly not get flushed to the database.

On systems with Swarming enabled, this could cause problems when a hosting agent tried to retrieve the highest alarm ID from the database.

Improve this page