A few minor bug fixes:

• Corrected a bug that may cause agent down alerts to not be sent for down agents with backup agents configured
• Corrected an issue in the Windows agent that caused the installer to run the service as user “.\”
• Corrected an issue in the Windows agent uninstaller that caused uninstallation to fail if the agent was installed in a non-standard location
• Corrected an issue that was preventing Windows event log alerts from being visible in the Alert tab
• An issue affecting the efficacy of enabling/disabling alerts at the top of the host tree was corrected.

A minor update was released March 8:

• Users with the Acknowledge role for a set of hosts can now schedule downtime for those hosts. (If you can respond to alerts, you should be able to prevent them….)
• Fixed an issue where editing roles would inadvertently remove the privileges for Services
• the remote SSH session functionality is now supported for any non-windows host (although there is still an outstanding issue with some Cisco devices)
• Fixed an issue where the Oracle active discovery was not using the supplied authentication credentials

A few minor bug fixes going out tonight:

• Alerts are no longer escalated to user accounts that have been suspended
• widening the legend field in graph definition UI
• getAlerts RPC call now supports StartEpoch and endEpoch to filter the returned set of alerts
• the RPC call to display a single widget no longer displays the trash/edit/delete icons on the widget
• a bug involving time based escalation chains incorrectly routing alerts on weekends has been fixed

Release notes for the February, 2012 LogicMonitor release, which will be applied to servers over the next week or so. (Account administrators will be notified about the applicable update time.)

New Features

• Alerts can now be routed to phones via Voice alert text to speech functionality.  Define a phone number of a LogicMonitor user, and add the contact destination user.voice to a stage in an alert chain, and the alert will be escalated to the phone number.  Voice calls are by default limited to 3 calls per hour per number. The first five calls a month are free. Further calls are charged – $25 for up to 25 calls.
• Alerts can also be delivered by WebHooks – POST (or GET request) to your servers when alerts occur.
• Support for Remote Sessions to monitored hosts – this feature allows you to securely forward SSH or RDP sessions from any web browser to monitored hosts. Communications are encrypted over SSL from the browser to the agent, then use the native protocol (SSH or RDP) to the end host. This ability is subject to access controls, and by default is disabled for all users. (Known issue: on Windows, username/password failures or RDP being disabled on the end machine both result in an Exception message.)

Improvements

• Map improvements: the host DisplayName is shown on the map with the location pin; improved color coding of location icons; map is displayed in the same window (to avoid pop-up blockers);
• Addition of a Map Widget for dashboards, showing the state of host groups by location
• Security:
• Ability to limit access to a LogicMonitor portal via IP address whitelisting
• An account is locked for 5 minutes after 5 failed login attempts.
• NOC widgets are now sortable (by status or group name)
• The system will prompt you to enter the reasons for a datasource update, so the history of changes are easily visible.
• Each dashboard widget now supports individually configurable refresh timers.
• The Alerts tab can now be set to automatically refresh, with a user defined refresh interval
• Ability to download SmartGraphs as PNG images
• A new schedule option for Active Discovery: run AD only when the host is initially added, or when Forced Active Discovery is selected. (Useful to prevent checking for port based datasources every day.)
• name-value pairs can now be separated by by colons, not just equal signs.
• Network Scan policies are now automatically saved when executed.
• Script datasources can now run powershell scripts natively
• External Website Monitoring/Services:
• Services alerts now identify which URL failed when following a chain of redirects
• The ability to schedule downtime from a specific collection point for Services
• Services search field – now supports searching for a partial url
• Support for HEAD and POST HTTP methods, as well as GET
• Support for custom headers
• ability to invert a match
• It is now possible to use host properties as virtual datapoints in a graph. (.e.g defining a property COMMITLEVEL and graphing a virtual datapoint ##COMMITLEVEL##) (This does not yet work on custom graphs.)
• improved xen collector performance
• The Link to URL field for a host now supports ssh urls such as ssh://192.168.1.1  (although a URL handler needs to be present in the OS or browser)
• Batch jobs now support scheduled down time.
• extended the supported length of comments for Scheduled Downtime
• Change to the format of alert Acknowledgement subject lines, to make them easier to distinguish from alerts. Acknowledgement messages are now of the form “Ack of alert by User“, rather than “Alert alert has been acknowledged by User“
• It is now possible to manually specify the color for aggregated datapoints on flexible custom dashboard graphs
• It is now possible to import new SNMP OIDs from the LogicMonitor core repository.
• The ability to match Windows informational level events in eventsources, providing the event ID is specified.
• inclusion of a DNS library in the agent for use in DNS server checks and more
• Agents now use persistent connections to reduce the number of connections from agent to our servers – this will reduce the load on proxies where the agent communicates through proxies.
• snmp can now be collected on ports other than the standard port 161, by setting the property snmp.port globally, per group or per host.
• The JDBC collector for database queries now supports a raw datapoint ‘status’, which will make it easier to distinguish between connectivity failures and authentication failures.
• An alert cleared message is now sent after an agent that was down has recovered.
• Logwatcher improvements:
• LogWatcher alerts now include the filer name and line number in the alert message
• the ability to support multiple patterns in one match tag
• You can now define exclude patterns to prevent lines that would otherwise match from triggering events.

UI Improvements

• Stages are now labelled with their stage number in the escalation chain editing form, and the editing buttons are now consistent with the rest of the UI.
• alert throttling is now referred to as alert rate limiting
• The datasource page will now prompt to save if there are unsaved changes when the user attempts to leave the page
• Default Min to 0 on overview graph creation
• “Show raw data” now shows the  values for calculated datapoints as well as direct datapoints.
• There is now a warning displayed in the alert if agent down alert not set
• Scheduled Downtime currently in effect is highlighted in red in the Manage Scheduled Downtime form
• Improved way to monitored Windows processes and services – right click …More..Add Monitored Resource
• Enforcing the inability to have groups names contain the reserved character ‘/’

Bugs corrected

• Clear alerts were incorrectly being sent if the alert cleared during SDT
• Host-down alerts were incorrectly sent when the issue was in fact agent connectivity issues, in some cases. The host down detection now requires positive messages from the agent.
• Various UI inconsistencies between the right click options and More button have been fixed
• Various spelling and other minor issues corrected.
• The Alert Filter on the alert tab was cleared in the UI after collapse/expand, but was still in effect.
• Hosts in SDT were not shown in blue on the google map display.
• Comments on creating SDTs for BatchJobs were not saved
• SDTs that crossed midnight were not always respected
• Acknowledged alerts were being unacknowledged in some edge cases
• Form based datapoint definition prevented the entry of time based thresholds.
• Dynamic Groups based on system.categories inherited by OID matching were not working
• It was not possible to delete manually added instances
• Corrected an issue preventing group names having spaces from being edited/created.
• Updated alert notes were not saved
• No longer require a double slash ( \\ ) to separate domain\user when specifying agent credentials for a proxy
• Added support for ##GROUP## as well as ##GROUPS## in alert templates
• corrected the inability to erase a property from being inherited by setting the property to blank at a higher level.
• Corrected an issue whereby the enforcing of strong SSL ciphers in agent/server communications prevented agent communicating via proxies.
• Services:
• an error prevented adding web services monitoring graphs to dashboards.
• Alerts were not treating the services folder as a group for alert routing
• Alert cleared messages were sent even if disabled
• Incorrectly followed redirects when “Follow Redirection” was set to No

Known Issues:

• Modifying a windows event source filter does not correctly cause the agent to apply the new filter until another configuration change is made
• There may be an issue with time based escalation chains with a single stage and alerts being routed to it outside of the defined time – under investigation

Release notes for the December, 2011 LogicMonitor release, which will be applied to servers over the next week or so. (Account administrators will be notified about the applicable update time.)

New Features

• Google map integration – you can show where hosts are located on a map that reflects their status.
• It is now possible to monitor the same IP address (e.g. 192.168.1.1) in two locations, so long as they are assigned to different agents and have a different display name.
• SNMP v3 support for those sites requiring encrypted SNMP communication
• Cluster Alerts: it is now possible to alert on the overall status of hosts or instances in a group, rather than individual hosts.

Enhancements

• Event sources now support more than one active alert at a time per eventsource, allowing individual event acknowledgements, etc.
• Event sources now have finer grained control, allowing disabling of an event source, or disabling the alerting of the event source, per host or per group, as well as showing the effective alert routing, and the ability to reply to Event alerts with the usual actions (SDT, ACK, etc.)
• Allows the logo for the login screen to updated from the Settings..Account page
• The host search field now includes the host description in the searched text.
• The alert tab can now be sorted by group column
• Adding a note to an alert no longer automatically acknowledges the alert.
• New users are emailed on account creation.
• Hyphens are now allowed in usernames
• You can now either delete a non-empty host group, leaving it’s member hosts, or delete the group and all hosts.
• Immediate execution of a network scan now allows the specification of a group for newly discovered hosts to be automatically placed into.
• Ability to pass arguments into netscan scripts
• Addition of a ##GROUPS## token in alert messages, which is replaced with the list of groups a host is a member of.
• A rewrite of the Netapp API collector to deal with performance issues in OnTap 8
• NetApp object discovery using the OnTap API now supports filtering by any object attribute
• Ability to choose multiple groups in alert, inventory reports
• New report class – Website Service overview
• Website Services now have a pseudo datasource of “website”, for purposes of alert routing and filtering.
• Web site services now support an Advanced monitor, allowing multi-step web transactions for synthetic monitoring
• External Alerting via an agent sending traps now support configuring the snmp trap community.
• Using a script on an agent to externally process alerts is now more flexible, as the standard set of tokens can now be passed to the script as parameters.
• Long polling interval – the ability to collect data as infrequently as once per day
• The webpage collector now supports circular redirects, up to a maximum of 3 redirects.
• The ability to specify arbitrary ports when using HTTP Active Discover method and regular expression capture groups.

UI Improvements

• graph tooltip now formats data to K, M, G
• Chart tooltips now show datapoint description
• Smarter legend wrapping for charts
• Addition of checkboxes allowing the enabling/disabling of monitoring or alerting for all instances at once, on the Manage Instances and Alerting form.
• Metric Trend reports now support multiple host groups
• Full screen widgets (suitable for embedding in an iframe)
• A variety of improvements to web site services – the ability to Show Alert Routing, Show Raw Data, Show Alert History

Bugs fixed

• graphing a non stack after stack was incorrectly treated as stacked element
• multiple “Alert Cleared” emails were incorrectly sent to CC recipients
• Metric trend reports sorted by Metric value now are correctly ordered (by the average for the reported period.)
• enabling time based escalation chains would clear previous stage definitions
•  disabling alert on a group level did not disable alert on subgroups within that group
•  maximizing NOC dashboard widget resulted in “page not found”. It now takes you to the Alerts tab.
• In some edge cases, Host Down alerts were incorrectly sent after agent declared down. The agent being down should suppress host down alerts for hosts the agent monitors.
• Under some cases, a user with view access to a limited set of hosts could create reports about hosts the user did not have role based access to
• Errors acknowledging snmp trap event alerts
• The Invert Match flag on website services monitoring was incorrectly set in all cases
• The requirement that group names be unique has been correctly changed to only require a unique name within the same parent group.
• time-based escalation chains and scheduled downtime schedules can now extend across midnight (e.g. support a  time range like 18:00 – 8:00)
• editing a group with empty properties defined (to block the inheritence of the properties from higher in the host tree) would incorrectly remove the blank properties.
• It is now possible to specify the datapoint type (gauge, counter, etc) on a script datasource datapoint if using the output directly.

Known Issues

• There is no visual indication if event sources are disabled on a host or group level.
• Clicking the link of a cluster alert in the alert tab will generate an error.
• if a device is replaced with a different kind of device with the same IP address, the system.sysinfo property is not updated.

Release notes for the September, 2011 LogicMonitor release, which will be applied to servers over the next week or so. (Account administrators will be notified about the applicable update time.)

New Features

• Datasources that support multiple-instances (wildcard datasources) are no longer identified by the name ending in a hyphen (which was just something you had to know, without any UI hint), but by whether the multi-instance checkbox is selected in the UI.
• integrated billing: LogicMonitor now supports integrated billing, which gives us much more flexibility in pricing – no more adding hosts only in blocks of 25.  We’ll be in touch to discuss your pricing options, and whether you want to stay on your current pricing scheme. (Entirely up to you.)
• new graphing engine.  We’ve adopted a java script graphing engine, replacing the Flash based previous version.  This means: no need for flash; better graph resolution; drag-to-select for zooming; native graph support on iphone/ipad, etc.
• external website monitoring open beta.  The Services tab will allow you to monitor your web sites from multiple locations around the internet – not just that of your agent. This will be a for-fee service, but it’s free for now (and also a bit unpolished.)
• Revamped network discovery scan system.  Discovery policies can now be associated with multiple agents, and agents can be associated with multiple policies.
• Composite Datapoints are now treated as regular datapoints. “So what?” you say.  This means they can be graphed, and reported on, and used in flexible dashboard graphs directly, instead of having to be re-defined in the graph, or report, or dashboard graph.
• Datapoints are now defined via a form based editor, which provides much greater guidance.
• Alert Reports can now present summarized report counts, so it’s easy to see the source of alert noise.
• Groovy scripts and postprocessor can now access system properties directly: hostProps.get(“system.hostname”);
• Datasource audits: you can now verify any differences between LogicMonitor core datasources and your own datasources, and the system will mark the datasource as audited until the datasource is updated at LogicMonitor core.
• Instance thresholds are kept if an instance is removed in discovery, then discovered again within 30 days.
• Availability of predefined Windows event log sources for IIS, SQL server, general system monitoring. (Exchange events are coming soon.)
• Active Discovery is automatically suspended during SDTs
• IPMI severity levels automatically mapped to LM severity
• Addition of filterSDT parameter to GetAlerts API to only show alerts not in SDT.
• ability to access JMX calls from agent scripts via groovy
• Addition of join() and exists() LMS functions
• Ability to access all host properties as alert tokens (except group membership – coming next release.)

UI Changes

• Ack icons changed
• Improved windows event display
• dropped support for “scale by 1024″. This is better dealt with in graphs by calculating scale in the graph. (ie. dividing by 1024/1024/1024 to covert to GB, etc)

New Datasources

• Mongodb
• APC air conditioning units
• Juniper and Extreme switches and routers
• Fortinet firewalls
• Zookeeper

Bugs Fixed

• Make Ports list in AD_jmx support same formats as AD_port
• user-agent “LogicMonitor” not used in active discovery
• clean up Windows Event Log filter options, and several bugs in event log filtering
• Dynamic host groups involving functions not refreshed correctly
• The incorrect agent was shown in the UI in a mouseover after an agent failover

These are the release notes for the August, 2011 LogicMonitor release, which will be applied to servers over the next week or so. (Account administrators will be notified about the applicable update time.)

Note: one UI change that may stump some people – the Datasources tab has been moved to an item off the Settings tab.

New Features

• First release of graphical reporting: you can now embed a graph per host, or summarizing multiple hosts, in a metric trend report.  More graphical reports are coming soon.
• Company information – accessible under Settings…Company Information and Timezone, or by clicking the company name in the top right, above your login name. Subscribe to release notes and maintenance notifications, and optionally upload your logo. As part of the upgrade, LogicMonitor will update your account with the current contacts – please check to ensure these are current.
• addition of a MongoDB collector and datasource
• Logfile monitoring: you can now deploy the LogicMonitor logwatcher program on windows and linux hosts, and use it to monitor application or OS log files, triggering alerts when certain strings are found.
• You can now disable debug functionality of specific agents entirely, which may be necessary for specific agents to comply with certain security policies.
• Host control of batch jobs: Ability to disable batch job alerting, disable the batch job, or SDT a batch job from the tool icon next to the batch job on the host view.
• Acknowledging an alert of a higher severity now applies to lower severity alerts,  if the status drops to a lower severity without first clearing.
• JSON and XPATH post processors to easily interpret data from JSON or XML web pages or script output. Details on this page.
• Ability to see datasources differences between the LogicMonitor repository and the local datasource when importing and updating datasources.
• Ability to put in comments when creating SDTs
• responseTime available to script collector
• Support ##DSIDESCRIPTION## in flexible custom graphs
• ability to specify a sort order for the host inventory report
• data downloaded from the smart graph is now automatically given a unique name, of the form host-graph_name-instance-time_range_str.csv

UI Changes:

• the alert level filtering on the Alerts tab search and the alert widget now show alerts of severity equal to or greater than the severity selected.
• datasources  now moved to the settings tab
• Alert messages will now refer to the datasource in alert by the Display name as used in the web UI (e.g. MySQL Global Stats), not the object name (e.g. MySQL-)
• Datapoint descriptions are now shown as tooltips when mouseover the object in alert settings on an instance
• Improved display of IPMI and syslog events
• Clicking an alert link no longer forces a Set Time Range operation. Time ranges have to be explicitly set.

Bug Fixes

• graph dropping to zero at edges should be resolved
• group column from alert tab now shows all groups a host is a member of, instead of just one.
• a variety of display issues in IE 9 now resolved
• overly long timeout on jdbc collector could cause some missing data due to blocking other threads
• readonly users can no longer change the layout of dashboards
• thresholds applied to groups were not propagating to hosts inside sub-groups.

• The ability to disable datasources at the group level.  If there is a set of datasources you wish to disable on a whole set of hosts, you no longer have to disable them on each host after Active Discovery has found them, or by changing the datasource Applies To field. You can now disable entire datasources on the group level.
• Top N dashboard graphs:  Custom graphs on dashboards now have a checkbox that can be selected to show only the top 10 objects.
• There’s been a variety of enhancements to reporting: cleaner report layouts, easier ability to select multiple hosts in reports, easier ability to select report datapoints, the ability to clone reports, and more.
• New report type – host inventory reports
• ability to search for hosts by ip addresses or partial IPs: the search box now finds hosts by IP addresses, or partial IP addresses, even if the host is added by DNS name.
• Scheduled Down Time (SDT) is now supported for eventsources (traps, event logs, IPMI events and syslog.)
• The Manage SDT forms now show who created each SDT.
• It’s now much easier to monitor a specific web page or UNC path on a host: simply use the Monitor a Web Page or Monitor a UNC path to add a resource to monitor on a host.
• Role Based access control now supports disabling the “Chat with Engineer” button for specific roles – this means MSPs can still enable access to the support button to connect with LogicMonitor support staff for internal usage, but not have the button displayed for their customers.
• The alert tab now contains a column showing a group the host is in. (Currently, only one group is shown for hosts that are members of multiple groups.)

UI Changes

• “Destination macros” have been renamed to “Recipients lists”
• The number of hosts in each group is shown in the tooltip for that group
• The first part of any alert acknowledgement comment is now visible in the alert panel
• All SDT dialogs now contain a “Quick Add SDT” button, which adds a one time SDT from the current time, for one hour.

Bug fixes

• “forgot password” link fixed
• sdt notifications were not sent to all alert recipients when alert suppressed in the UI, as opposed to replying to the alert via email/sms.
• fixed the appearance of custom graphs dropping to zero on the right hand time line.
• fixed some drag and drop errors that previously required a page reload to correct.

New Datasources

Monitoring now covers Asterisk, SIP, pfSense, Force10, Brocade lb, Cisco switch stacks, Redis, Pick D3 databases and more.

And as always, if there is hardware or software that LogicMonitor does not cover – just ask!

These are the release notes for the March, 2011 LogicMonitor release, which will be applied to servers this week. (Account administrators will be notified on the applicable update time).

New Features

• syslog event monitoring
• extension to the groovy toolset, allowing easy scripting of arbitrary HTTP transactions.
• the ability to change the time range of regular host graphs when shown on a dashboard. (Click the pencil icon on the widget, select a new time frame, and submit)
• Wizard to aid in constructing Applies To fields. When defining datasources or dynamic groups, you can now use a Wizard to construct the LMS expression that controls which hosts will be associated with the group or datasource.
• The CC field is always notified when a new SDT is created. (Previously, it was an optional checkbox)
• Significant improvements in Windows Event Log monitoring – cleaner display, autocomplete for event source filters
• You can now select hosts by group, as well as host name pattern matching, for SLA reports, metric trend reports
• The comments included in a reply to an alert, that also creates a SDT, will now be included in the notification message sent to those to whom the alert has been sent to. (e.g. if an alert is replied to with “SDT 7 will check in morning”, then the “will check in morning” comment will be included in the message sent to others that the alert had been sent to.)
• enhancements to the embedded Expect client.
• The ability to disable the “Chat with Engineer” button on an account wide basis. (Currently you have to contact support to disable the button. Role base control will be coming soon.)
• Expanded tokens available for event alert messages
• More information is automatically collected about ESX hosts (CPUs, hardware model, etc)
• Added the ability to reinstall an existing agent

Other Changes

The http collector uses a different library in this release. The significance of this is that reported HTTP page time now includes the connection setup time – previously, that time was not reported.  This will only significantly affect results for long latency measurements – for example, those page response times being collected by an agent several thousand miles away will show an increase in reported page load time.

UI Changes

• The Suppress button (which scheduled SDT for the instance in alert starting at the present moment) is now part of the Acknowledge form.
• Comments now required in acking an alert.
• simplified log in screen – the company field, which was always pre-filled, is now hidden entirely
• changed “Test alert” icon to “Show alert routing”, to clarify that it does not trigger alerts, but simply shows the routing alerts will be subject to.
• snmp interface descriptions no longer repeat the interface name before the description
• agents are now referenced by their description, not just numeric ID, when reporting agent down, or when confirming agent deletion
• ESX, JMX, memcached, Perfmon and Xen specific datasources have had the UI definition greatly simplified. (Other collectors will have the same improvements applied next release)
• The duration for Scheduled Downtime is now calculated and shown when start and end times are used.
• when viewing a host, the System information (the OS version, display name, and other discovered properties) are now separated into the System tab, while user set properties are in the “Customized Properties” tab.

Bugs Fixed

• corrected an issue where tasks with long periods between collections would sometimes not be scheduled correctly if other collecting tasks changed frequently.
• The relative ordering (Display Priority) of overview graphs was not respected.
• Search results were broken on the Reports tab
• special characters (e.g. “/?! ) now display correctly in graph titles.
• sms template was not used, even though a user had SMS format paging preference defined.

These are the release notes for the January, 2011 LogicMonitor release, which will be applied to servers this week.

New Features:

• ESX  hardware monitoring – ESX hardware is now detected and monitored automatically, using the CIM providers VMWare provides for common server hardware. Note that hardware monitoring is only displayed for ESX hosts if you add the ESX host itself to LogicMonitor – it cannot be monitored only by adding VCenter.
• IPMI discovery, monitoring and event logs are now supported. This provides a method for monitoring hardware status for devices that do not expose other hardware monitoring methods.
• Dashboard management has been rewritten to be more powerful and intuitive.  The Add Dashboard, Delete Dashboard and Manage Widgets buttons have been replaced by a single Manage Dashboards button, that takes you to a screen that allows you to add/delete/edit dashboards for individual users, or shared dashboards.  From this screen you can also create/edit/modify widgets for any dashboard, or copy and move widgets. See here for more details.
• The Show Properties/Show Alerts button in the hosts view has been replaced with a tabbed view.
• Agent installation has been further simplified. The agent credentials and configuration are now included in the binary downloaded from the server, so no configuration is needed.  Note: one impact of this change is that you cannot re-use the same agent download to install on more than one host.
• Ability to compare metrics for hosts within a group.  When clicking a group in the Host view, all the datasources of all members of the group are shown, as before. However, when expanded, a list of hosts with instances of that datasource are shown. You can then easily expand each host, to allow comparison of the datasource across members of the group. (more…)