- General Enhancements and Fixes: Enhanced Kubernetes monitoring, addition of new Middle East AWS region, new properties for website monitoring, and more.
- New and Updated Monitoring: New LogicModules have been released for the monitoring of AWS Kinesis Video Streams, GCP Cloud Trace, Microsoft Windows Failover Clusters, Palo Alto, MongoDB, Silver Peak, and more.
- Entering Beta: Push Metrics API.
IMPORTANT: Beginning with the rollout of v.146, LogicMonitor is updating its new version deployment window to 7:30-10:30 p.m. PT (previously, the window was 6-9 p.m PT). We will continue to provide in-app notification 48 hours in advance of a new version deployment.
General Enhancements and Fixes
- Ability to link alert to its associated components. The alert detail pane now provides hyperlinks to the resource/website/service associated with the alert and the DataSource (if applicable).
- (FIX) Alert acknowledgement. Resolved an issue that was causing exceptions when users with a particular combination of permissions attempted to acknowledge alerts.
- (FIX) Cleared ConfigSource alerts. Fixed issue preventing cleared alerts triggered by ConfigSources from displaying in the Alerts page.
- (FIX) Time zones in graph legends. Resolved issue where graph legends (as displayed on the Graphs tab of the alert detail pane) were not honoring the user’s configured time zone.
- (FIX) Severity icon display. Resolved an issue preventing the display of severity icons in some cases when viewing the pop-out version of an alert from the Alerts page.
- AWS updates:
- AWS Bahrain region added. Added support for the AWS Bahrain region (me-south-1), which is the first AWS region in the Middle East. Customers can now monitor services in the Bahrain region that are supported by LogicMonitor’s AWS monitoring.
- New StorageType property for AWS RDS. Added StorageType as a property for AWS Relational Database Service (RDS). This new property (system.aws.storageType) allows logical grouping and filtering according to StorageType.
- (FIX) AWS Kinesis Streams permissions. Previously, AWS Kinesis Streams permissions were not being properly tested. In some cases, this resulted in customers not being alerted that additional permissions were required to be able to monitor AWS Kinesis Streams from the LogicMonitor portal. This issue has been resolved.
- Alert List widget. The behavior of the anomaly filter available in the alert list widget configurations now aligns with that of the anomaly filter available on the Alerts page.
- (FIX) Custom Graph widget. The Custom Graph widget was returning an “invalid virtual datapoint expression” error when attempting to calculate virtual datapoints that referenced datapoints belonging to the same DataSource, but applying to different resources. This issue has been resolved. If you previously experienced this issue, open the Custom Graph widget configuration dialog and simply save the current configurations – no edits are necessary.
- (FIX) Rounding of Y-axis values. In some cases, the Y-axis value for some graphs was not displaying as a rounded value. This issue has largely been resolved, although there is still a remaining corner case that we are actively working on where non-rounded values still display if display type is selected as “Line” and data is constant (data is displayed in a straight line) for virtual datapoints
The following new features and improvements will be available with the new v5.0.0 version of Argus, which is targeted for arrival in early January.
- Avoiding name conflicts in discovery. To avoid name conflicts and subsequent issues with discovery, the following enhancements were made:
- Resource type (for example, “Pod”) is appended by default to the LogicMontior display name
- New Argus Helm chart options, FullDisplayNameIncludeNamespace and FullDisplayNameIncludeClusterName, are available to append namespace and cluster name respectively to the LogicMonitor display name (default: false). You may find that enabling one of these two new options is useful for avoiding naming conflicts in cases where pods with the same names are running across multiple namespaces and/or clusters.
- For resources that fail to be added due to name conflicts, the conflicting resource will be added to a dynamic group named “_conflicts”.
- Discovery filtering by pod name. Pod name is now available for use with discovery filtering, including handling of special characters such as hyphens. You may find this useful for filtering pods in/out of monitoring based on names rather than labels.
- Discovering EKS Fargate nodes and pods. Previously, nodes and pods for EKS Fargate weren’t being discovered properly due to shared IPs, but no host network specification in the pod manifests (EKS Fargate pods each run on a distinct node and share IPs, but don’t use the Kubernetes Host Network specification). To account for this, we’re now categorizing pods with the EKS Fargate label “eks.amazonaws.com/fargate-profile” as pods configured with host networking. This will ensure (1) the right value for the system.hostname property gets set, (2) the right DataSources apply, and, most importantly, (3) there are no discovery conflicts between pods and nodes for EKS Fargate.
- New LogicModule statuses. LogicModule statuses “Beta” and “Deprecated” now display in the LogicModule metadata table.
- Log events +/- three hours no longer accepted. Log events with timestamps more than three hours in the past or future are no longer accepted, and will result in a new 4006 API response code. This is to ensure relevance of data and protect against misconfigured timestamp settings.
Network Traffic Flow Monitoring
- New benchmarks. New capacity benchmarks for network traffic flow have been published for the Large, XL, and XXL Collectors. These benchmarks show a dramatic increase in the number of flows per second that can be comfortably monitored and were reached as a result of an ongoing initiative to enhance and optimize LogicMonitor’s network traffic flow monitoring capabilities. See Collector Capacity for the new benchmarks.
- Support for sFlow version 5. Added support for sFlow version 5 (Collector version 29.105 or higher is required).
- DNS lookups. Increased the DNS cache storage duration to 24 hours (previously duration was 30 minutes) to resolve repeated DNS queries preventing the resolved names from displaying.
- (FIX) IPv6 address format display. Normalized IPv6 address format display for NetFlow IPv6 endpoints that do not resolve via DNS.
- (FIX) Locked API tokens. Previously, locked API tokens due to multiple unsuccessful authentication attempts were causing an exception to occur. This issue has been resolved.
- (FIX) Updating a device via PATCH request. Previously, submitting a PATCH request to modify both a device “name” and its customProperties would result in the device name updating to the name of one of the customProperties. This issue has been resolved.
- (FIX) /apiStats/externalApis. Previously, /apiStats/externalApis was returning separate values for API calls to the same endpoint depending on whether the user made the call with or without a trailing forward slash (/). This has been corrected and calls with either variation are now appropriately identified and aggregated as a single entry in the response.
- New Web Check properties. Several new properties are now available for manipulating the settings of Web Checks. Specifically, these settings provide some flexibility in cases where sites may load slower than the default timeouts used by LogicMonitor’s external checkpoints.
The following built-in properties all default to 30 seconds and all support an acceptable value range between
>0 and <=60.
- website.http.client.so.timeoutInSec (socket timeout in seconds)
- website.http.client.connect.timeoutInSec (TCP connection timeout in seconds)
- website.http.client.read.timeoutInSec (read timeout in seconds)
New, Updated, and Deprecated LogicModules
Recently released to the LogicMonitor public repository, our new and updated LogicModules are available for import to expand and enhance your monitoring coverage.
|(DataSource) Gathers throughput, latency, request and error metrics for Kinesis Video Streams.
|(DataSource) This new DataSource deprecates the now-legacy snmpFreeBSDDisk- DataSource. Enhancements include an updated AppliesTo statement that includes OpenBSD systems; updated filters to exclude /proc and /dev and allow discovery of files between 0-100 bytes on filesystems; new complex datapoints; and support for drives over 8 TB and block sizes other than 4 K.
Note: See the Deprecated LogicModules section of these release notes for details on how to transition from a deprecated to a replacement LogicModule.
|(DataSource) Monitors billing spans (billing/monthly_spans_ingested and billing/spans_ingested) during the polling interval and month-to-date for Cloud Trace.
|(PropertySource) Retrieves additional Juniper device information for virtual chassis members where PropertySource Device_BasicInfo is unable to do so.
|(DataSource) Monitors Kubernetes Horizontal Pod Autoscaling (HPA) metrics related to health, scaling status, replica counts, system stats, latency and Grafana requests.
Note: If you are using an existing version of Argus (4.2.0 or a lower version), you may not see HPA resources discovered due to insufficient permission and may need to upgrade to 4.3.0 (releasing shortly) first.
Note: The addCategory_NoPing PropertySource has been updated to apply the category of “NoPing” to monitored Kubernetes HPA devices.
|(DataSource) A new DataSource for monitoring latency of the Collector buffer data reporter (via JMX) has been released to specifically work with Collector versions 29.105 and higher. The existing DataSource (LogicMonitor_Collector_BufferDataReporterLatency) is still required for versions 29.104 and lower.
Note: LogicMonitor recommends that, even if you’re not yet using Collector version 29.105, you still import this new DataSource. The newer DataSource will not associate until the Collector version is updated to 29.105 or higher at which time it will automatically associate, ensuring a smooth monitoring transition. At that time, the older DataSource will cease to function.
|(12 DataSources) New monitoring for Microsoft Windows Failover Clusters that features new/improved monitoring for cluster state, disk utilization, MRR messages, message/network throughput, reconnection rate, message queue size, interface state, individual/summary node states, node vote weight, node count, print server jobs, and RCM metrics. These DataSources apply to the cluster VNN rather than each node individually. For setup and monitoring details, see Windows Server Failover Cluster Monitoring.
Note: Seven DataSources are deprecated by the release of these new modules. See the Deprecated LogicModules section of these release notes for details on which DataSources were deprecated and how to transition from a deprecated to a replacement LogicModule.
|(3 DataSources, 1 PropertySource) With the release of these new LogicModules, LogicMonitor is supporting two sets of MongoDB LogicModules simultaneously. Both sets offer the same coverage, but this new set uses a newer driver that is required for compatibility with Collector version 29.106 (and higher). Because of a hard JAR dependency, the older set only works with Collector versions under 29.106. For setup and monitoring details for either set, see MongoDB Monitoring.
|(18 DataSources, 3 ConfigSources, 1 PropertySource) A large number of new LogicModules have been added to the Palo Alto firewall monitoring package to significantly expand monitoring capabilities. New monitoring includes certificate expiry, actions taken by filter policies, licensing, log forwarding queues, zone active connections and various 24 hour report metrics.
Note: As highlighted in the Updated LogicModules section of these release notes, LogicMonitor has also updated 24 existing Palo Alto firewall monitoring LogicModules.
|(5 DataSources, 2 PropertySources, 1 TopologySource) Monitors disk and memory usage, next WAN hops, tunnel states, and overall system status for Silver Peak devices. Includes topology support. For setup and details, see Silver Peak Monitoring.
Note: This new monitoring package was originally released in v.144, but because it was inadvertently left out of the initial publishing of the release notes for v.144, it is being included here to ensure awareness of its availability.
|(2 DataSources) Designed to eventually replace snmpHRDisk- and snmpHRLargeDisk-, these two new DataSources offer superior monitoring. Status monitoring has been pulled out from usage monitoring to prevent filtering of instances where filesystem usage data would be desired. The Percentage used is no longer adjusted by 5% for root reservation, and a separate datapoint has been added to account for root reservation.
Note: It is intended that these two new DataSources will eventually replace snmpHRDisk- and snmpHRLargeDisk-. However, due to widespread usage, we will not be deprecating these at this time and, instead, are rolling out the new DataSources without alert thresholds in place. Recommended alert thresholds are documented in the technical notes if you should choose to enable alerting.
|(PropertySource) Updated collection script to fall back to using the “/public/cluster” endpoint to determine version (previous endpoint was known to fail on Cohesity version 6.5.1).
|(PropertySource) Updated to apply the category of “NoPing” to newly-monitored Kubernetes HPA devices (as enabled by the newly-released Kubernetes_HPA DataSource).
|(PropertySource) Updated to use here-strings to allow for credentials containing special characters.
|(DataSource) Updated technical notes to indicate that BadBattPacks is no longer reported in XL devices after firmware 5.1.7.
|(ConfigSource) Filter edout uptime value from output.
|Azure Advisor Recommendations
|(EventSource) Updated alerting to map severity from Azure Advisor.
|(DataSource) Converted collection method to scripted and added logic to retry collection multiple times within the interval to prevent data gaps; added datapoint to measure the number of retries needed to successfully collect per interval; renamed Temp datapoint to Temperature, which will cause historical data to be lost for the old datapoint.
|(DataSource) Replaced invalid characters in wildvalues; fixed collection output.
Note: Some instances may be deleted upon upgrade, however, these instances should have not been collecting data due to the issues addressed by the upgrade.
|(DataSource) Fixed invalid and duplicate wildvalue/alias being set; updated collection output; increased SNMP timeout on both scripts.
|(2 DataSources) Re-implemented caching layer; added code to gracefully handle credential changes to ensure that ‘loggedout’ tokens are refreshed.
|(DataSource) Added a two-second delay after sending “show time” command to resolve collection issue; updated technical notes.
|(EventSource) Fixed invalid token in debug messages; increased loopback timeout.
|(DataSource) Changed the collection interval to match the Active Discovery interval of 15 minutes; updated to alert after two consecutive failures to prevent excessive alerting.
|(DataSource) Fixed issue where replacement of invalid characters was causing collection to fail upon empty metric return.
|(11 DataSources, 1 PropertySource) Removed generic catch statement that was obscuring error messages; fixed incorrectly scoped HTTP object that was preventing sessions from being closed correctly on failure; updated collection scripts to ensure all available metrics are outputted; removed the numOfMappedScsiInitiators datapoint from the EMC_ScaleIO_Volumes DataSource.
|(DataSource) Updated Active Discovery script to no longer discover non-Unity datastores.
Note: Upon update, non-Unity datastore instances will be deleted.
|(DataSource) Updated the OID being used by the peerStatus datapoint (previous OID was incorrect).
|(DataSource) Updated disk parsing regex to allow for “?” characters and extended state data.
|(DataSource) Updated datapoint descriptions; removed alert from SizeOfBigQueue datapoint; fixed errors in the names of the SyslogMessagesIgnoredDueToHostNameRetreival and SyslogMessagesIgnoredDueToHostNameRetreivalRaw datapoints (name changes will result in the loss of historical data for these datapoints).
|(3 DataSources) Added alert to indicate incompatibility with Collector versions 29.106 and higher. As discussed in the New LogicModules section of these release notes, a new set of MongoDB LogicModules has been released to work with Collector versions 29.106 and higher.
|(DataSource) Fixed expression for TotalArrayUsagePercent complex datapoint; updated expression used in the collection script to match “snapshot usage” rather than “uncompressed snapshot usage” to support data printout formatting used by newer versions of Nimble software.
|(DataSource) Updated scripts to align with Nimble_HDD_Disks DataSource; added support for multiple arrays; added disk serial as instance description; updated tech notes and graphs.
|(23 DataSources, 1 ConfigSource) For DataSources, converted to batchscript where possible; moved to using builtin groovy URL library; Added instance-level properties where relevant; updated technical notes.
For ConfigSource, updated collection script to no longer use deprecated HTTP library; fixed output formatting to allow collected configs to be restored without modification. Note: Due to the nature of the formatting change, all devices that the ConfigSource is applied to will generate an alert upon update. Note: As highlighted in the New LogicModules section of these release notes, LogicMonitor has also released 22 new LogicModules for significant enhancement of Palo Alto firewall monitoring capabilities.
|(DataSource) Changed graph to show raw usage instead of percentage usage. Updated alert message for StorageNotAccessible datapoint.
|(ConfigSource) Add regex to exclude “ftp password” from configs.
|(DataSource) Fix incorrect post processor parameter for the GuestToolsVersionStatus datapoint.
LogicMonitor deprecates LogicModules for a variety of reasons. Most often, it is because we have published a replacement LogicModule that provides superior access to the monitored technology (for example, supports more versions of the technology or provides improved scalability).
However, there are times when a LogicModule is deprecated with no replacement (for example, when a technology becomes defunct or security issues beyond LogicMonitor’s control arise).
Note: When importing a replacement LogicModule, you will not experience any immediate data loss due to the name variation that LogicMonitor expressly adds. However, there will be a diversion in data collection between the deprecated and new LogicModule, and you will potentially collect duplicate data and receive duplicate alerts for as long as both LogicModules are active. For this reason, we recommend that you disable monitoring of the DataSource instances at the resource or resource group level after you have imported its replacement. When DataSource monitoring is disabled in this way, it stops querying the host and generating alerts, but maintains all historical data. At some point in time, you may want to delete the legacy DataSource altogether, but consider this move carefully as all historical data will be lost upon deletion. For more information on disabling DataSource monitoring, see Disabling Monitoring for a DataSource or Instance.
|End of Support Date
LogicModules Deprecated in v.146
|(1 deprecated DataSource) Enhancements featured by replacement include an updated AppliesTo statement that includes OpenBSD systems; updated filters to exclude /proc and /dev and allow discovery of files between 0-100 bytes on filesystems; new complex datapoints; and support for drives over 8 TB and block sizes other than 4 K.
|(7 deprecated DataSources) The replacement DataSources apply to the cluster VNN rather than each node individually, significantly reducing the incidence of duplicate alerts.
Note: The six replacement DataSources listed here are part of a bigger monitoring package for Microsoft Windows Failover Clusters that was released in v.146.
LogicModules Previously Deprecated (over past five versions)
|(1 DataSource) It has been determined that all data gathered by this DataSource is being reported more robustly through the existing NetSNMPCPUwithCores and NetSNMP_Memory_Usage DataSources, making it redundant and unnecessary.
|(4 DataSources)General cleanup of DataSources that are not widely used and no longer work correctly.
|(1 DataSource) Converted to a scripted DataSource and uses an additional OID to resolve duplicate wildvalues.
Liebert Power lines
Liebert Temperature probes-
Liebert UPS Temperature probes-
|(3 DataSources) Rewrote DataSources as scripted SNMP; merged monitoring of temperature and UPS temperature probes into a single DataSource; added new power and voltage metrics for power lines monitoring; fixed an incorrect filter for power lines.
|(1 DataSource) LogicMonitor released five new DataSources that provide 15 additional datapoints and group related metrics into separate DataSources, making it easier to hone in on specific aspects of your portal. For more information, see LogicMonitor Portal Monitoring.
|(1 DataSource) Active Discovery and data collection methods were converted from SNMP to script to improve overall efficiency.
|(1 DataSource) “VMware” was removed from the DataSource name as the backups are not VMWare specific.
|(1 DataSource) Windows SNMP support has been deprecated by Microsoft since Server 2012 R2. As a result, this DataSource does not collect reliable data for Server 2016 and above.
Cisco AIP SSM Memory Pools-
|(1 DataSource) The replacement DataSources offer a scripted SNMP version that returns accurate values. If you were previously using the now-deprecated DataSource, it can be deleted as it was providing invalid data.
|(1 DataSource) Uses a deprecated API. It is strongly recommended that this LogicModule be deleted and the replacement LogicModule be used.
|(1 DataSource) The replacement DataSources filter out instances to allow for separate DataSource monitoring of memory, flash storage, and standard storage.
|(1 DataSource) Did not provide accurate/useful data and can be deleted; the existing Linux_SSH_ServiceStatus DataSource has been retooled in v.141 to perform the same function.
|(1 DataSource) The replacement DataSource offers several advantages including (1) the use of the more widely supported nstat (part of iproute2) command and netstat as fallback if nstat is not available and (2) additional monitoring for packet, datagram and error related metrics
|(1 DataSource) Changed the collection method from standard SNMP to scripted SNMP to properly handle the units returned for the VolFreeSize and VolTotalSize datapoints. In the now-deprecated version, these datapoint calculations were failing if the returned values were not expressed as terabytes leading to data misrepresentation.
Entering Beta: Push Metrics API
LogicMonitor is developing a Push Metric API in order to allow metrics to be sent directly to LogicMonitor without the use of a Collector. Some technology landscapes where this capability is becoming vital include server-less infrastructures such as AWS Lambda, large distributed microservices environments with custom metrics (including business metrics), and IoT environments.
Beta for Push Metrics API is targeted for January. If you have a compelling use case and the resource availability, and would like to be considered for participation, please complete the Customer Beta survey.