GROUNDWORK-1848

GroundWork Messenger test button

GROUNDWORK-1718

Data Geyser should suppress most duplicate data for check results

GROUNDWORK-1510

Update downtime automation scripting for GroundWork 8

GROUNDWORK-1019

Add alias and notes fields to Status

GROUNDWORK-1775

Add hostgroup memberships to host detail in Status

GROUNDWORK-1895

TCG Kubernetes layout minor bugs

The Kubernetes connector configuration screen was a bit messy, we cleaned it up

GROUNDWORK-1890

Prometheus remote write not working with GroundWork APM connecter

There were issues pushing data to the APM connector with some Prometheus configurations, we enabled them

GROUNDWORK-1889

Kubernetes connector config by file broken

Some config files were not getting parsed correctly, leading to auth failures in Kubernetes, now corrected 

GROUNDWORK-1886

SLA dashboard doesn't load if there are empty servicegroups

The SLA dashboard had issues loading when empty service groups were selected, now loads with no issues 

GROUNDWORK-1869

BSM prefetching of hosts no longer working

A handy feature of BSM that made the first run after a new configuration action more complete and accurate was broken, now fixed

GROUNDWORK-1855

tcg-elastic fails to retrieve metrics with logged error

There was a bug which could impede the collection of metrics from Elasticsearch in some cases, now fixed 

GROUNDWORK-1853

tcg-kubernetes saves config to volume only when configured via file-upload

There was a serious problem loading configuration files for tcg-kubernetes, which made this method of auth unstable past restarts, now fixed 

GROUNDWORK-1843

status-queue worker exception behavior needs optimization

Part of an optimization effort that makes Nagios data (and Parent Child communication) much faster

GROUNDWORK-1802

Elasticsearch and Logstash have a vulnerable log4j version

A non-exploitable yet vulnerable version of jog4j was installed on GroundWork 8.2.2, this was removed, and scans now show it as clear

GROUNDWORK-1900

Notification method cURL allows shell commands within GroundWork GORT container

A security issue if used improperly, this has been inhibited

GROUNDWORK-1899

SLA dashboards with custom groups no longer work

An issue with SLA dashboards that affected custom groups was corrected

Nagios Monitoring

Renaming a host to the same name except for case will require 2 commits. The renaming of a host to the same name with differences only in the case of the characters is possible, but requires committing the configuration twice.

Run the Commit (or, for Parent Managed Child, the Build Instance) process a second time, and the renamed host will appear. 

Nagios Monitoring

Adding two hosts with the same name except for case will not work. If you ever find you need to have two hosts with the same name except for case in the Nagios system, this is allowed. It will not show both hosts, however, only the first one you add, and it won't work at all if you do it in a single commit.

To do this (admittedly really rare thing), fist add one of the hosts. Choose the one with the name you want to see, add it and commit. Then add the second and commit. The second host's services will be merged into the first. 

LDAP

Cannot delete users created by LDAP sync. If you sync LDAP to GroundWork, the users who log in are imported. There's not much point in deleting them unless you disconnect the LDAP server, but even if you do so, the delete function isn't working.

There is a work-around, but then again there's no harm in leaving these user accounts alone. They will fail login once the LDAP server is inactive. If you really must delete them, please contact support. 

Nagios Container

It's possible to overload the Nagios instance in GroundWork. Nagios uses workers that communicate with the services it runs. In some cases the default provisioning isn't enough to keep up with a really active alert stream.

You can override the default provisioning of workers in Nagios with the check_workers directive. See How to configure check_workers.

Custom Reports

Optional Jasper Server slow to start. If one starts the optional Jasper Server, it may take several minutes to become available.

Wait several minutes after starting the system with the Jasper Server loaded before accessing the Custom Reports menu item. 

TCG-SNMP restart generates unknown status

When restarting the (new) TCG-SNMP connector, the initial poll may generate UNKNOWN status for the network interfaces. 

We recommend adjusting the notification rules for this to avoid alarm fatigue. 

Cron-Controlled GroundWork containers generate SUSPENDED alerts

Jobs scheduled VIA the cron container generate SUSPENDED alerts. This is correct behavior, but may cause unwanted alarms.

We recommend adjusting the notification rules for this to avoid alarm fatigue. A separate rule for the DOCK-H:cadvisor group with a 1-minute delay can suppress these alarms.

Unattended mode install is not enabled

The installer won't support the --unattended  mode option to install standalone with default settings in this version. 

This convenience feature will be enabled again in future versions.  

Archive Migration initially incomplete

Upgrades to the system will migrate the foundation database but will not completely migrate the archive database until the first archive run (overnight in the cron timezone)

This may affect any queries you want to run against the archive database, especially those for host address values, but only for the day of upgrade.