It's time for This Week in OpenNMS!
Sorry I missed last week, but I was out for a bit and by the time I was back figured I'd just punt to this week.
In the last two weeks we did more work on moving code to using OIA, improved flow persistence, fixed a bunch of alarm bugs, worked on a bunch of code debt, did more topology UI work, and made lots of improvements to Helm.
Github Project Updates
- Internals, APIs, and Documentation
- Chandra continued his work on service detector support in OIA.
- Dustin worked on making it possible to use JSR-223 scripting to modify flow data before it is persisted.
- David has been working on adding explicit priority support to events for more predictable ordering.
- Matt worked on being able to send inventory directly to in-JVM OCE without using Kafka.
- Christian added support for transforming nodeId and interface in JEXL when using the
event:stress
command. - Jesse worked on optimizing alarm (re)publishing to occur only if
lastModified
orackTime
have changed. - Christian updated the DHCP service monitor to use dhcp4java.
- Chandra did more work on service monitor support in OIA.
- Jesse continued develpment indexing alarm history in Elasticsearch.
- Dustin worked on tracking flows by application rather than source and destination port.
- Jesse worked on an example set of Drools alarm rules to create "nag" notification events.
- Matt added correlation feedback support to OIA.
- Matt did more work on fixing support for additional syslog formats.
- Christian fixed a bug where
linkDown
alarms would not be generated ififIndex
is unknown. - Matt worked on improving performance of alarm snapshotting in the Drools context.
- Patrick started on the thankless job of modernizing the SnmpCollector. (Thanks!)
- Patrick worked on normalizing a number of places where we convert integers from strings.
- Jesse did a number of fixes to improve edge cases and failures in the new alarm handling.
- Web & UI
- I did a ton of enhancements to Helm (mostly in the alarm table code) in preparation for a 3.0 release.
- Christian added applications to edges in the BSM topology view.
- Antonio continued his work on a topology ReST service.
- Dustin added support for using the browser notification API for getting notifications.
- Dustin fixed the measurements API to return responses in the same column order as requested.
- Patrick worked on wrapping up his topology performance enhancements.
- Sebastian continued to work on the Bootstrap 4 UI refactor.
OpenNMS Discourse
As many of you are surely aware, traffic on our e-mail discussion lists has steadily declined for a long time. People have, for the most part, been moving to other venues for asking questions, and a lot of folks have embraced https://chat.opennms.com/ as a way to get real-time help from other community members.
Thanks to the fine folks at Discourse, we now have a community forum that can hopefully meet the needs of folks who traditionally used the e-mail lists. I realize it's no replacement for the oldskool listservs I still hold in my heart, but it at least provides a lot of tools for notifications and replying by e-mail to give some amount of equivalence to the mailing lists.
For more details, please read Ronny's post about updating our community spaces, and if you want to dive right in, go to https://opennms.discourse.group.
OpenNMS Holiday Break and Releases
The OpenNMS Group will be on break next week for the holidays, and we will resume normal operations in the new year.
Also, a note: normally we'd do releases on the third Thursday of the month, but this time around that is... the day before we disappear for a week. So obviously that's not the greatest idea, and we'll be skipping this month's releases. The next Horizon and Meridian update scheduled is Thursday, January 17th.
I hope everyone has a happy holiday and a good new year! See you in 2019!
Until Next Week…
If there’s anything you’d like me to talk about in a future TWiO, or you just have a comment or criticism you’d like to share, don’t hesitate to say hi.
- Ben
Resolved Issues Since Last TWiO
- HELM-98: Alarm Table text bold if 'Style with severity' is checked
- HELM-115: Drag'n Drop doesn't work with Safari
- HELM-116: Alarm table no longer respects column sizing
- HELM-117: Unable to deselect a row in the Alarm Table
- HELM-118: Enable Table panel to display durations
- HELM-120: Sorting by number of alarms in the alarm table panel
- HELM-121: Related alarm view should show the node label
- HELM-123: Drag&drop-reordering of columns broken in Firefox 63 and Internet Explorer 11
- HELM-124: Limit number of alarms returned by the FM DS
- HELM-125: Add the ability to see the complete alarm JSON from the alarm details view
- HELM-126: white text when Style With Severity=Row is hard to read
- HZN-1437: Cannot clear alarms
- HZN-1440: Create tool to audit syslog message parsing
- HZN-1449: Improve performance of DroolsAlarmContext alarm snapshot handling
- NMS-8622: jmx config generator test failing
- NMS-10337: Replaced jdhcp with dhcp4java
- NMS-10405: Allow Applications to be added as an edge to a Business Service
- NMS-10439: Apply initial loading improvements to IsIs, lldp, ospf protocols
- NMS-10459: Meassure and improve performance of Interface loading and mapping
- NMS-10461: Alarm list pagination broken with situations
- NMS-10473: duplicate key exceptions in alarmd
- NMS-10475: When using the events:stress command, the node-id or interface passed as parameters are ignored when using jexl
- NMS-10477: Use a utility function instead of Integer.getInteger
- NMS-10479: Alarm Dashlet CriteriaBuilder In-Restriction not working
- NMS-10480: Standard output get wasted with "Unknown record type: 0 / 2209" messages
- NMS-10483: Make favicons in the web application great again
- NMS-10487: Meassure and improve performance of Cdp/Lldp/IsIsElement loading
- NMS-10494: Drools alarm may update alarms unnecessarily on snapshot
- NMS-10502: Create Application Icon for BSM Topology
- OIA-2: Add support for ServiceDetectors