Get OpenNMS
Get Involved
Get to Know Us
From OpenNMS
Contents |
Download
- Released versions
- Nightly snapshots (YUM and URPMI, ZIP file at end of list for all OS-es)
Stable Releases
New in OpenNMS 1.8.3
Bug Fixes
- Performance in the outage editor has been improved. (bug #1225)
- The "move up" and "move down" buttons in group modification were broken. (bug #4034)
- Remote pollers cause excessive event logging. (bug #4045)
- OpenNMS logging was still set to DEBUG, even though we're in a stable cycle. (bug #4047)
New in OpenNMS 1.8.2
New Features and Enhancements
- It is now possible to schedule an outage for a node or interface directly from their respective UIs. (bug #1225)
- Linkd is now able to walk IpCidrRouteTable and InetCidrRouteTable as well as IpRouteTable for determining node links. (bug #3773)
- You can now configure the number of outages to show on the front page of the web UI. (bug #3982)
- A standalone tool for importing Spectrum traps is available in the opennms-tools directory of the OpenNMS source. (bug #4025)
- A standalone tool for parsing collection logs and providing stats on average collection time is available in the opennms-tools directory of the OpenNMS source. (bug #4033)
- It is now possible to pause and resume all remote pollers in the distributed poller status UI. (bug #4026)
- You can now populate asset data based on information retrieved from SNMP when provisioning a node. (bug #4028)
- New events and configurations for Junos devices were added. (bug #4029)
- Queued now gathers RRD queue statistics by default. (bug #4030)
Bug Fixes
- The thresholding for low memory now takes cached memory into account on Linux. (bug #2910)
- Editing a notification's name in the GUI now works again. (bug #3570)
- The custom resource performance report link now gives you a custom resource performance report, instead of a standard one. ;) (bug #3573)
- A resource filter with a negative look behind now renders properly in the UI. (bug #3726)
- Browser history buttons now work properly in maps again. (bug #3734)
- A bug in linkd that would prevent many nodes to fail to have proper links was fixed. (bug #3989)
- A regression from 1.6 where the "nodes with outages" box on the front page would show inconsistent results was fixed. (bug #3981)
- A bug was fixed in the code for editing node labels where it would previously cause a database handle leak (and eventually, errors accessing the database) in OpenNMS. (bug #3991)
- If you had a "." in a node label, maps would not populate their name properly. (bug #3995)
- A bug where saving maps could cause duplicates to be created was fixed. (bug #3997)
- A typo in the RANCID property used in opennms.properties was fixed. Please double-check your configuration after upgrading. (bug #4004)
- In some cases, code that handles new nodes could generate an inconsistent NodeGainedService event. (bug #4022)
- The distributed poller status legend needed clarification. (bug #4027)
New in OpenNMS 1.8.1
New Features and Enhancements
- Back by popular demand, it is again possible to show the number of events in the event listing. To enable it, edit opennms.properties. (bug #3675)
- The quick search box was moved to be below the current outages box on the front page. (bug #3849)
- Events for Net-SNMP diskTable DISMAN traps were added. (bug #3850)
- Query-by-example now lets you filter by node in the notification list. (bug #3899)
- The remote poller map's Mapquest implementation now supports click and double-click to mimic Mapquest's behaviour. (bug #3901)
- Next and previous links are now at the bottom of event, notification, and outage list pages, as well as the top. (bug #3910 and bug #3932)
- You can now use %shorttime% in notifications for a more compact date. (bug #3917)
- A vacuumd automation has been added to automatically clean out old remote poller database entries. (bug #3923)
- Errors at startup relating to sockets (address already in use, etc.) will now appear in the manager.log alongside other startup errors to ease debugging. (bug #3940)
Bug Fixes
- Fix problems on system with incorrect IPv6 configuration. (bug #1959)
- Fixed incorrect "The database server's error messages are not in English" errors on installation. (bug #3296)
- The "Add Node" UI is now smarter about what it writes to disk. (bug #3443)
- Sorting was fixed on the advanced alarm search results. (bug #3589)
- The opennms-contrib debian package now properly depends on libxml-twig-perl. (bug #3722)
- The Solaris packages play better with system boot now. (bug #3802)
- The Debian and RPM packages are more explicit about what needs to be done on post-installation. (bug #3848)
- The data collection failed event tells you what failed again. (bug #3869)
- A number of null pointer issues that could cause linkd to not create links between nodes were fixed. (bug #3871)
- Spring security settings were fixed so dashboard-only users can properly connect to the dashboard again. (bug #3900)
- Inline thresholding is enabled by default again, as it was late in the 1.6.x series. (bug #3912)
- A number of regex pattern validation bugs in our config handling have been fixed. (bug #3915)
- The notification browser now properly shows only a given user's notifications when a filter has been applied. (bug #3939)
- The jar installer now handles errors a little more gracefully on Windows. (bug #3953)
- Syslogd had a number of line-matching bugs that have been fixed. (bug #3958)
New in OpenNMS 1.8
This represents what is new in the upcoming OpenNMS 1.8 series.
Architectural Changes
- DAO Updates
- Many more parts of the OpenNMS data access layer have been updated to be compatible with Hibernate and database abstraction.
- Database Versioning
- The OpenNMS upgrade process now uses LiquiBase for managing schema changes between versions.
- OSGi Preparation
- Some daemon architectural changes have happened to support an eventual integration with OSGi (post-1.8).
- Spring 3.0
- OpenNMS was updated to use Spring 3.0.
- Test Framework Updates
- A number of handy annotation-based test framework changes have been made, using JUnit 4.
New Features
Core and Infrastructure
- Pinger
- A new multithreaded, high-performance ping infrastructure was added used by all subsystems that talk ICMP.
- RADIUS Authentication
- A RADIUS authentication provider has been added.
- RANCID Integration
- An integration with RANCID has been added. If configured, extra data about node configuration, etc. will be available in the node UI and maps.
- RESTful Interface
- A number of interfaces to OpenNMS data have been made available through a RESTful interface, using the Jersey API.
This includes simple web API access to alarms, events, nodes, notifications, and outages, and it is expected that this will be expanded in future releases.
Discovery and Capability Scanning
- Provisioning
- A complete replacement for Capsd, called "Provisiond" is introduced in this release. It allows you to define specific behaviors for detecting services and attributes of devices in manual, automated, and semi-automated ways, with simple APIs for writing your own custom scanning behavior and detectors. This includes a highly-scalable, highly-parallelizable threading architecture which will be used for other parts of OpenNMS in future releases.
- DNS Importing
- Provisiond is able to provision nodes based on a DNS zone transfer.
- Automatic Map Creation
- Provisiond can automatically create maps and link nodes together when they are provisioned.
- New Capsd Plugins
- JDBCQueryPlugin, JMXSecurePlugin, WebPlugin, WmiPlugin, XmpPlugin
Monitoring and Data Collection
- Mobile Sequence and SMS Ping Monitors
- A new pair of monitors have been added that have the ability to perform monitoring of cellular networks with phones attached to the OpenNMS server.
- Remote Monitor
- The remote monitor has been enhanced to communicate with the OpenNMS server through HTTP (which makes it proxyable).
- TCP RRD Strategy
- A data collection strategy that allows pushing collected data over the network.
- SNMP Poller
- A service that let you snmp poll the interface operational and administration status (up/down) for selected interfaces
- WMI Support
- Support has been added for polling and datacollection from Windows Management Instrumentation.
- XMP Support
- Support has been added for the XML Management Protocol (Cartographer).
- New Poller Monitors
- BSFMonitor, JDBCQueryMonitor, JMXSecureMonitor, MemcachedMonitor, WebMonitor, WmiMonitor, XmpMonitor
Events, Alarms, Notifications, and Acknowledgement
- Acknowledgement Daemon
- A daemon for interacting with acknowledgements programmatically has been added. This will open up APIs for acknowledging alarms through cell phone, IM, e-mail, and so on.
- Asterisk Notifications Support
- It is now possible to integrate with an Asterisk server to send notifications as phone calls.
- Microblog Notifications Support
- It is now possible to send notifications through Twitter, Identi.ca, and similar services.
User Interface and Reporting
- Adobe AIR Client
- An Adobe AIR based client was added.
- SVG Map Updates
- The topological map support has been updated to work properly in all major browsers with SVG support, and has had a significant number of other enhancements.
- Remote Poller Maps
- A new map displaying remote poller status was added.
- Mobile Browser Cleanups
- Some updates have been made to the web UI to make them more mobile-friendly.
- Reporting
- Reporting has been enhanced greatly to support Jasper reports and better scheduling.
- Web UI Updates
- A number of pages, most notably the node details page, have been updated to provide more detailed information thanks to the new REST service. These controls also support paging, sorting, etc.
- Web UI Filters
- Support for basic access filters has been added, which allows you to limit which nodes, interfaces, and services are presented to a user based on the surveillance categories their group is assigned to.
