We continue to charge ahead with behind the scenes enhancements but have found the time to bring you some much anticipated features and improvements.  The May release (v42) will be rolled to customers during the second half of the month.

New Features

• NetApp SnapMirrors can be monitored by using the NetApp API.  When adding a datasource for a NetApp collector,  selecting API call as the collect method, snapmirror-get-status as the API field, and snapmirror-status as the Index Property will enable monitoring.  Parameters and datapoints must be added.

• A new report type, Interface Bandwidth, provides visibility to average and maximum inbound and outbound traffic.
• Services alerts can be configured based on multiple testing locations, enabling alerts to be refined based on all, half, or one testing location.  Status graphing is available for all testing locations combined as well as each individual location.

Improvements

Reports:

•  Date / Time data in Alert Reports is in the format yyyy-mm-dd hh:mm:ss to enable international date formatting.
•  Web Service Overview reports will display changes in status, treating status changes as events rather than performance data.
•  When creating an Alert report, the option “Include pre-existing alerts?” will include all alerts that meet the report parameters independent of the last activity date.  For example, an active alert report for the past day will only include alerts created and active with the last 24 hours.  Selecting ‘Include pre-existing alerts?’ will display any alert that has had a status of active in the last 24 hours.

Datasources / Event Sources:

• Simplified process to add Windows Services, Windows Processes, and Linux Processes as monitored resources, providing a real time list of processes to select from.  To add, right click on the host name, select More | Add Monitored Resource.
• New active discovery filter, packetsRX, supported for the ESX Host Interface datasource
• When creating or editing Log File EventSource, users were prompted for Alert Severity.   The severity is handled within the Log File Collector Attributes so the Alert Severity configuration has been removed.

Usability:

• The Alert History table has a new column for ‘Acked On.’
• Graph images are generated server side as pngs, improving processing times.

Bug fixes

• Various graphing updates, including an increased margin for the Y axis to accomodate large values, text wrapping in legends, and tool tips displaying datapoint descriptions consistently
• The drop down list for available groups when adding a filter in the Alert Tab has been reactivated.
• Collectors were rejecting a feed if a filter had been created by API that did not have a value, or a host without a hostname.  With this release, missing values will not cause an exception.
• Scheduled Down Time for event sources and data sources will be reflected in the Host | Managed Scheduled Down Time | SDTs affecting this object window.
• Invalid date formats no longer cause issues for collectors when processing syslog messages
• The Host group inventory group reports would include sub groups regardless of the include subgroups setting
• Fixed situations in which ##ALERTTYPE##, ##COLLECTORDESC## and ##COLLECTORID## tokens were not rendered correctly.
• Disabling a datasource at the group level was not updating child groups.
• The escalation level of an alert was being passed through when time based subchains were in use, causing a newly triggered chain to begin with an escalation point.  Time based subchains will now always start at the first stage.
• The application was not correctly displaying event source status in regards to Scheduled Down Time.
• Corrected various links to help documentation throughout site
• When an EventSource Log File Collector Attribute matched a trigger and a not trigger parameter the alert was sent.  Alerts will only be sent if the trigger matches and the do not trigger does not match.
• Host Cluster alerts were applying the thresholds of the specific hosts in place of the Host Cluster alert thresholds and the Host Custer Alert email was not being sent.
•  Ability for Readonly/ackonly roles to add hostgroups has been removed.
• When a new user was created and “Force Password Change” selected, the user was not prompted to change the password at next login.
• The application was not correctly displaying an acknowledged collector down alert
• Scheduled Down Time for hosts and instance groups could not be edited if the SDT review window was opened from right clicking Group Name  and selecting Show hosts with scheduled down time.
• Adding arbitrary emails to a stage in an escalation chain required commas to separate multiple addresses
• MsSql JDBC Active Discovery was not working as expected

The v41 release will be rolled out to customers over the next two weeks.

New Features

Alerting:
To enhance alert details and custom alerts we have added the following tokens:

• For all Alerts:
  ##HOSTDESCRIPTION## – Text description of host
• Agent Alerts:
  ##AGENTID##  - ID associated with the collector
  ##AGENT_DESCRIPTION## – Text description of collector
  ##BACKUPAGENTID## – ID associated with the backup collector
  ##BACKUPAGENT_DESC## – Text description of backup collector

New alert templates have been added to ensure critical information is included in alert messaging. Custom alerts will not be impacted by these changes. The new templates include:

• Service Alerts – Triggered by a service value based on checkpoint and start time
  
• Agent Down Alerts – Triggered if LogicMonitor has not received any data from a collector
• Agent Failover Alerts – Triggered if a collector fails over to a backup collector
• Agent Throttled Alerts – Triggered if the number of alerts exceeds the threshold within a set period of time

The generic alert clear template has been replaced with a per template default.  All cleared alerts will include the original alert subject and body.

• EMail and SMS template:
  Subject:  ***Cleared*** <original subject>
  Body:  *** This condition has been cleared
  <original alert body>
• Voice Messages:
  LogicMonitor reports this alert cleared.  <original alert message>

DataSources:

ESX VM Snapshots are now available as a DataSource. Monitor the age and size of VMware snapshots as well as VMs with the oldest snapshots and those consuming the most space.

User Management:

• Administrators can logoff all users from Settings | Roles and Users
  
• Administrators can force users to update passwords, for additional security and audit compliance.  The force password change checkbox is available from Settings | Roles and Users | Edit User.  When selected the user will be logged off and forced to enter a new password at login.
• Role changes now displayed in access log

Reporting:

• Generate JPEGs directly with the getGraphImage API call

Improvements

• Enable deletion of groups with host members, hosts no longer assigned to a group will display under the main company node
• Alert severity filter incorporated into NOC widget
• Drop down displays for country code when adding a user’s phone number
• Ability to set the content type for custom HTTP alerts to form data, URL encoded form data, JSON, XML or plain text
• Datasources that used the reserved word ‘time’ have had the datapoint name updated to ‘responseTime’.
• Improved XenServer monitoring performance, added Xenserver 6.0 compatibility
• Re-enabled ability to update your billing information under settings | account information
• Retrieve all paths for a host using getHost or getHosts
• Test alerts available for the email custom delivery method, webhooks and voice calls
• Added in VMWare snapshot monitoring datasource. Import from core.logicmonitor.com

Bugs Corrected

• ConnectWise integration was hanging on authentication for customers with a large number of Service Boards
• More corrections of typo’s, better input validation for forms, more informative error messages, etc
• Event Sources – Reason for update not required in all cases
• A variety of improvements in SiteMonitor, especially where checking JSON content
• Corrected an issue that caused multiple batch jobs in a group to be disabled incorrectly
• Voice call alert acknowledgements no longer add “Confirmed ” in comments
• Corrected an issue that caused MongoDB active discovery method to fail if authentication was required
• Host group inventory will only represent each host once regardless of group and subgroup membership
• Host group inventory reports will retain wildcards for host group filters
• Alert acknowledgement were being incorrectly applied to previous occurrences of the same alert
• Group level user privileges now apply to all member hosts.  Privileges are not affected by the host being a member of other groups even if the user has no privileges for the other groups.
• Complex datapoints were sometimes calculated incorrectly when there was no data on a component datapoint
• SubGroup alert thresholds were not being applied correctly if Group thresholds existed
• Added ability to re-enable ‘View and ack alerts for *’ for ackonly role
• “Manage Scheduled Downtime” from top right button does not function
• Dashboards require users to have host permissions, if a user does not have rights to view a host included in a dashboard widget they will receive a permission denied error
• Corrected an error that caused metric trend reports to fail with an exception occasionally
• Corrected a bug that caused alerts to not respect the alert clear transition interval in all cases.
• Site monitor now supports checking sites with 256 bit encryption
• Corrected an issue that prevented the disabling of monitoring or alerting at host group level

Minor release for Feb 28, 2013, which will be applied during the standard maintenance window.

New Feature

• There is now the ability to set a users View Permission. This controls which tabs are visible to the user in the UI. This setting overrides settings in Roles. (e.g. a user associated with a role with permission to view hosts can have the ability to view hosts removed by clearing this permission.)  Currently this feature applies only to users – it will be added to Roles at a later date.

Improvements

• A new datasource available to be imported - VMware resource pool monitoring.
• Extension to the API getGraphData, so that custom graph data can be downloaded
• A new token ##STARTEPOCH## is available for use in alert templates, which is replaced with the time (in unix epoch time) when this alert started. Useful for creating unique alert identifiers.
• Account payment management buttons to add/update credit cards are temporarily removed, while we update our backend billing system. In the interim, please contact our Billing department to set up or change credit card payments: billing@logicmonitor.com 888 415 6442, select option 3

Bugs Corrected

• There was an issue with the release of the LogicMonitor proxies: where the collector was blocked from connection to the LogicMonitor application proxies by a transparent proxy at the customer’s location, and received (correctly) an incorrect SSL certificate, it blocked the collector from reporting data. This has been corrected, along with some log message discrepancies in switching to the application proxies.  Note: we are not pushing the collector upgrade during this minor release.  We believe we have identified and addressed any customers affected by this, but if you reconfigure your network so that your collector has to transit a transparent proxy that blocks the LogicMonitor proxies, you may be affected and should re-install the collector.
• Corrected an error that could lead to NetScan adding hosts with an empty hostname, which could lead to collector failures
• Adds the ability to SDT HOST via email for event alerts (while SDT replies worked for the event source, they did not support SDTing the entire host.)
• Corrected a bug that led to some bi-weekly scheduled reports being sent many of times
• Corrected an issue that led to the “No global thresholds” option not being able to be disabled on the Alert Threshold Report
• Corrected and issue that led to duplicate alerts in the alert report (from alerts tab)
• Corrected an issue that led to 64 bit Windows collectors being unable to use integrated authentication with Sql Server.
• Corrected an issue that led to tokens in alert templates being treated case sensitively, incorrectly.
• Corrected an issue with netflow data display incorrectly if the byte value is more than 1G.
• “Manage Scheduled Downtime” from the More.. button did not function
• A bug was introduced in the last release that prevented Alert Cleared messages for Collector Down alerts. This has been corrected.

There was a lot of behind the scenes work in this release, rolling out more systems and more metrics for us to monitor on our own infrastructure, but we still had time to add many things to improve LogicMonitor from our customer’s point of view. This release will be rolled out to customers over the next few weeks. Check out the video for the highlights.

New Features

• Integration
  • Looking for an easy way to integrate LogicMonitor Alerts into other systems? As well as webhooks, we now support email as a “Custom Delivery Method”, allowing alert messages to be sent with a consistent format, for easy communication with systems that require a fixed format of email.
  • Much simpler Connectwise Integration
• Internet Transit Reliability: Internet connectivity issues between collectors and LogicMonitor should now be effectively eliminated. LogicMonitor now runs proxies in different regions of the internet, providing improved server-collector communication even in the face of internet issues.
• We’ve laid the foundation in the collectors for an even wider variety of device monitoring – this collector upgrade now includes DB2 and SMI-S capability. Look for datasources utilizing these soon.

Improvements

• Scheduled Down Time:
  • Host alert table now includes a new column “InSDT” to clearly show whether the alert is in Scheduled Downtime.
  • It’s now easy to see all the inherited SDTs that apply to the object as well as those that impact it directly, with an improved way to view Scheduled Downtimes.
• How much has an alert cleared by? There is a new token available for use in the alert clear template, that will report the current value of the reported data: ##CLEARVALUE## We recommend you add the following line to your alert clear template (under Settings…Alerts..Other Alert Settings.. Default Templates):

  Reason for clear: ##DATAPOINT## is not ##THRESHOLD##: the current value is ##CLEARVALUE##

• Usability Improvements
  • Drag/drop to move groups to a different parent group.
  • Wizard to help construct complex and time specific alert thresholds
  • Dashboards no longer perform an unnecessary refresh when an absolute  time range is selected
  • Google Maps widget loading performance is now greatly improved (after it’s been loaded once), and the hover display has been improved.
  • Removed the Alert Level column from alert widget (as severity is conveyed by color, this frees more space for the alert text)
  • The NOC widget now sorts Acknowledged alerts before no alerts.
  • Smarter legend wrapping for Service status graph
  • An easy way to see all disabled alerts on a host: Show Disabled Alerts menu item
• Security Enhancements
  • Fixed an issue that allowed non-privileged user to see password properties if they inspected the JSON calls to the server within their browser
  • All passwords are now required to have a minimum of 5 characters
  • The audit log retention period is increased to 60 days (from 30 days)
  • Corrected a security issue where users with restricted host access could create private dashboards showing metrics about hosts they did not explicitly have access to by using wildcards.  Now, in order to view any information on a dashboard, the users must also have at least read access to the hosts in the host tab.
• Collector changes
  • collectors that are deleted within LogicMonitor now shut themselves down and uninstall themselves.
  • The collector installation wizard no longer automatically deletes unregistered collectors if the installation is cancelled. This is useful where the collector cannot communicate back to LogicMonitor until proxy settings are configured.
  • Support for native SMS alerts on collector down notifications
• Add the methods HTTP.body and HTTP.headers to the exposed groovy HTTP methods
• Website Services report now uses the same graph engine as other reports
• a new function is available in datapoint expressions to test whether a value is a member of a set of values: in(expr1, expr2, expr3, …, testval)
• New API calls: GetHostGroup by ID, verifyProperties, confirmAlerts, getAlerts enhanced to include the SDT status of the alert

Bugs Corrected

• Eventsources that had alerting disabled had no indication of that in the UI. Now they show with an Asterisk next to them, consistent with datasources.
• Deleting host groups also removes them from any Google Map widget
• Dashboard Auto-Rotate Functionality has been fixed to a per-user setting
• Corrected an issue where the conditions on a filter on the Alert tab would show inverted after a refresh (although functioned correctly)
• Corrected an issue that could cause collector restarts when the collector was monitoring a very large set of hosts with complex configurations, that caused the configuration file to take a long time to generate.
• Improved input validation and layout in many forms
• Corrected an issue where Cluster Alerts that were set to disable individual alerts only did so on direct members of the group, not sub-groups
• Corrected an issue with the Services monitor sometimes incorrectly resolving DNS names
• Corrected an issue with the jdbc data interpreter ignoring column aliases (select foo as blah…)
• Corrected an issue with event alerts not respecting datapoint filters in Alert widgets
• Corrected report footers showing incorrect timezone
• Corrected an error when running a Host Group Inventory Report on the property “Category”
• Updated datapoints with the reserved name Time to the name ResponseTime
• A restricted user with no RBAC rights to the settings tab could access (but not see or set) settings by clicking the Company link
• Corrected an error preventing editing of the Name of an overview graph widget
• Corrected an error preventing scrolling to the Submit button for the Manage Alerts and Thresholds form in IE with lower resolutions screens.
• Updated the “Show Associated Hosts” form to show hosts reporting Netflow data to a collector
• Corrected an issue with incorrect encoding of some active discovery parameters that could cause complex regular expression filters to fail
• To work around issue of Xenserver reporting the last known stats for powered down VM’s, discovery now checks the VM power state
• Fixed an issue with SiteMonitor not functioning correctly with long URLs
• PNG graph downloads were named incorrectly – they were truncated at first space
• A bug prevented viewing service folder scheduled down times.
• Cluster alert SMS templates were not correctly applied.
• Corrected an issue with Mongo DB collector not releasing threads appropriately.
• Corrected some edge case issues with WMI authentication.
• Hosts in subgroups were showing up twice in flexible graphs
• Hosts using SNMP v3 with authentication specified, but without privacy protocol, were not handled correctly
• Corrected an issue with Cluster alerts showing an incorrect group in the alert
• The Alert Threshold Report now shows hosts with Alerting disabled, and instances with all alerting disabled on the instance level. These were previously omitted.

These fixes will be applied to portals tonight, at around 9:00 pm PDT during the regularly scheduled maintenance window.  The only impact should be a few seconds of portal unavailability.

Bugs corrected

• Event sources using a filter of REGEXNOTMATCH and a property that was empty on the host level would incorrectly exclude all events.
• Failed script based Active Discovery tasks could incorrectly remove all instances if the script failed due to a resource issue on the collector.
• Corrected some issues  Sitemonitor scalability (this was actually corrected last night.)
• Corrected a few issues with new collector installation, regarding proxies and initial installation.

These fixes will be applied to portals tonight, at around 9:00 pm PDT during the regularly scheduled maintenance window.  The only impact should be a few seconds of portal unavailability, and the need to log in again.

Bugs Corrected

• the voice message for alerts clearing used incorrect verbiage, so the fact the alert was cleared was not apparent.
• The information in Settings/account information page was not being displayed correctly after an update
• There was a delay in displaying Services Alerts
• corrected an error in displaying Smart graphs with non-default time ranges
• Corrected an issue with incorrect validation of E164 phone number format for UK numbers
• Web Services could not deal with Services with characters such as & in the name.
• Native SMS was disabled for South Africa
• Collector installation wizard was not saving the escalation chain and interval
• Syslog event collector was not handling the MoreUrgentThan operator correctly

Improvements

• The Collector management screen now shows the number of hosts in the collector grid
•  A bigger multiline textbox to edit webhook payload

This update will start being rolled out to production customers this week.

With our increasing numbers of European customers, we have expanded our live support hours: 9 am London time all the way through to 5:30 pm Pacific, Monday – Friday.  We’ve also changed the chat icon, so that if no engineers are online, it allows you to leave a message, which opens a support ticket.

Improvements

• Dashboards
  • Added the ability to clone a dashboard.
  • It is now possible to download the image from graph widgets, directly from the dashboard, by means of the download icon on the widget.
  • “Full screen” mode for dashboards (click the green icon with 4 arrows to the right of the dashboard tool bar)
  • Flexible custom graphs can now have hosts selected by group membership, as well as host pattern matching
  • The event widget (which reported collector level events) has been removed – the functionality is available under the Settings..Collector mangement ..Check Collector Events screen.
• Reports
  • Prettier Reports! Graphs in reports now use the same graphing look as graphs in the web interface.
  • New report type, Alert Threshold Report, allows you to easily see all the thresholds in effect for a set of hosts, and if they have any thresholds overridden.
  • Ability to show 95th percentile lines in metric trend reports.
• Alerts and Alerting
  • Alert thresholds for a host can now be edited directly from the Alert panel for that host – click the pencil next to the alerts and you can see the global and group level thresholds, and edit the threshold for this instance.
  • Alert Cleared messages are now under template control. (Alerts are under Settings..Alert Settings..Other Alert Settings..Default Alert Templates.)
  • Scheduled Downtime can now be set on the Instance Group level (currently only for manually managed instance groups.)
  • Improvements in webhooks (many more tokens, consolidated UI, etc.)
  • Improvements in native SMS delivery
  • Explicit throttling for voice calls has been removed, and uses the escalation chain throttle controls.
  • Reverted to behavior that alert clears are sent to all contacts that received the alert, instead of just the acknowledger.
• Collector
  • The Collector configuration utility on Windows can now automatically detect proxy settings configured for Internet Explorer.
  • To avoid any issues with automatically updated versions of Java, collectors now install their own JRE.
  • Reduce collector’s use of abortive TCP connection release.
• Passwords now require minimum of 5 characters
• The Logwatcher has been retired as a standalone system, and it’s functionality is now merged into the collector
• Customizable New User message
• Updated the  ESX Host- datasource for all customers, so that the ESXi hosts CPU is changed from host.cpu.usage.average to host.cpu.utilization.average. The newer counter is more accurate with hyperthreading, so you may notice a slightly higher reported CPU usage.
• Sitemonitor (service checking) now supports Ping checks from the external locations, instead of just HTTP and HTTPS.
• External website checking now reports and graphs fullpage load if configured to do so.

Bugs Corrected

• There was an incorrect threshold set on the ESX Hardware Health- datasource for some customers. This has been corrected.
• Corrected an issue where group threshold were not applied in certain edge cases
• A Datasource with an invalid regular expression defined for a datapoint would prevent all other datapoints from collecting. This is corrected so that other datapoints are reported correctly, and an error logged in the collector logs.
• Corrected an issue where a datasource change that should have been non-impacting led to an inconsistency between server components internally, blocking data from being stored (but not affecting alerting.)
• Excludes Cisco IOS XE devices from incorrectly matching the IsLinux() function
• Corrected a bug where mysql- datasource did not timeout if the connection was accepted by the server, but not responded to
• Corrected an issue with JDBC active discovery where a combination of a long URL and a long SQL query would exceed the internal field length, and prevent the datasource being updated.
• Added scroll bars when needed to the “Show Hosts with Scheduled Downtime” option of the top node of the host tree, to accommodate displaying many hosts in SDT on smaller monitors.
• Corrected an AJAX caching issue that caused adding a new Service to prevent turning off monitoring or alerting for all services until a page refresh
• Corrected an issue preventing users with “Ack only” role for certain hosts from seeing Event alerts on those hosts.
• Corrected an issue that prevented “Remote Session” functionality via SSH from working with Cisco devices
• Users with Manage Hosts role could add and delete hosts, but not change thresholds on the hosts. This is corrected.
• Improvements in Perfmon data collection
• Corrected an issue whereby instances could be removed by JMX active discovery if the JMX service was not responding. It was corrected to only remove instances where the service responds, but reports the instance no longer present.
• Corrected an issue with recent versions of Chrome not showing PDF or CSV reports
• Datasources in a display group were not showing the form that allowed SDT to be defined on a datasource at a group level.
• When viewing script datasources in IE and chrome, sometimes they were intermittently unviewable in the datasource pane
• Corrected duplicate displaying of datasource and eventsource display groups.
• Corrected an issue with expanded dashboard graphs showing at the top of the page, even if the graph being viewed was located below the fold.
• Users were not able to rearrange the layout of widgets even on their private dashboards, unless they had administrative roles. Corrected.
• Corrected an issue involving time based escalation sub-chains being re-ordered on saving of time-based chain
• Corrected in issue in CSV report formatting, where incorrect escaping of characters such as commas resulted in incorrect tabulation.
• Corrected an issue where dashboard graphs downloaded via FireFox had their filename truncated to just the host
• The “Download Report” from the alerts tab was not being subjected to role based access control.
• Corrected an error with Oracle database monitoring when the collector was returning no data when authentication failed (as opposed to return “No Data”)
• The Tutorial dashboard videos were not playable in Chrome
• Corrected an issue where Cluster Alerts that were set to “Suppress Individual Alerts” were not suppressing individual alerts in sub-groups
• Corrected the time zone settings for UTC+02:00 Helsinki, Athens, Bucharest, Cairo, Pretoria, Sofia
• Alerts to SMS email addresses were no longer using SMS template if custom template set for datapoint
• SMS Alerts Short Template were incorrectly including verbage “Please click URL …”

A feature that made it into the last release, but did not make it into documentation – until now.

You can now use properties within event source filters.  This is a lot cooler than it sounds – it means you can now add one line to your event source filters, and use properties to customize what events match the event source. So you can set properties globally to exclude an event or events from all hosts; override that by setting the property per group; or override that on a host level.

Want to exclude event ID 123 from most hosts, have a group of hosts trigger on event ID 123 but not on 234 or 456, and a single host not trigger on any of these events?  You can now do all this with the same event source. Just set a FILTEREDEVENTS property on the top, host and group levels.

For more information, see the Host and Group level filtering section of the Windows Event Monitoring documentation.

Note: the property based filter will not be present in the event sources of current LogicMonitor accounts. (We avoid changing your monitoring once you are running.) You can just add in the following line to your event sources yourself, or contact support to have it done for you.

A few minor bugs squashed:

• Certain reports were showing content for all hosts where only certain hosts were selected.
• The Mongo datasources were using the properties mongo.user and mongo.pass instead of the documented properties mongodb.user and mongodb.pass. Now both sets of properties work.
• Corrected an issue where some mails systems (mostly Exchange) encoded the ACK message in a way that caused the reply not to be understood.
• Corrected an issue where reports displayed as HTML on recent versions of Chrome would display an error.  The requirements for the Content-Disposition header in Chrome changed – reports now conform with the new requirements and work correctly.

Bugs corrected

• Corrected a bug where some Service Alerts were cached incorrectly, leading to incorrect setting of the alert level during configuration.
• corrected an issue that resulted in some data not being returned to the server after the last upgrade, where the datasource had many datapoints
• Corrected a bug where the Set Password checkbox retained it’s state on other user edits, which could result in inadvertent changing of passwords
• Correct a bug where the report email delivery schedule was cached, and carried over to other reports created in same session
• Corrected an issue where some email encodings led to the ACK reply not working

Improvements

•  alert ID’s are now included in the message body of alerts, as well as the subject, so that replies from systems that strip out the subject can still successfully acknowledge the alerts. (e.g iPhone on Sprint PCS)
• Change to the default alert template for voice calls for events, to only read the first 10 words of the event
• support HTTP Basic Authentication for HTTP Alert Delivery (webhooks)
• EC2 network host scan can now be selected to use the EC2 internal IP or the public IP for each instance