...
- the NMIS installer now reports missing modules better, and aborts (by default) if there are any crucial missing modules
- NMIS now handles critical sections and long-running NMIS processes much better .than before
- PID files are no longer used, and instead the actual process table is checked.
- New config option
max_child_runtime
can be used to limit collect/update job runtime. - If NMIS jobs runs over-time, a warning event is generated unless config option
disable_nmis_process_events
is set to false. - Collect jobs can now optionally run longer than a single collect cycle, if the command line option
ignore_running=true
is given - in which case NMIS will just warn about old NMIS processes that are still running, not kill them as it does by default. - The NMIS Collect Runtime graph now includes average and maximum numbers number of NMIS processes.
...
- .
- NMIS now includes a full-fledged node administration tool for command line and script use
admin/node_admin.pl
can create, export, update and delete nodes- it can also rename nodes, and all RRD and other node database files are renamed accordingly
- nmis.pl can now add any groups that are set for specific hosts but not globally configured (command line argument
type=groupsync
).
This is not done automatically as groups are meant as an administrative mechanism for selecting/skipping nodes' in the GUI. - NMIS now consistently supports "true/1/yes" and "false/0/no" for setting boolean configuration values everywhere.
- 8.5.4 includes improved new versions of the support tool and the nmis_file_cleanup.sh data pruning tools
- logging was improved logging, reduced false positives in the logs
- improved device support for Net-SNMP-based hosts and Windows hosts, Cisco devices
- new and improved device support for Huawei routers, including Quality of Service collection and graphing
- improved rrd_tune_interfaces.pl helper script now sets interface speed maxima, and correctly adjusts them on repeat execution
- improved improved
fpingd.pl
with more precise logging. fpingd now can also optionally log stateless node information (see otion optionfastping_stateless_log
) for external consumption - event escalation is now handled more robustly and consistently
- improved activation of opMaps and opFlow display widgets
- improved file permission handling for improved performance
- numerous graph improvements and bug fixes
NMIS 8.5.2G
Tuesday 9 September 2014
...