* Added snmp mibs from vendor (https://www.hillstonenet.com)
* Added memoery discovery and pool
* Added serial and os version discovery.
* Added cpu discovery.
* Added temperature and power sensors
* Added test data.
* Fix yaml format.
* Pass mib name as parameter.
* Pass mib name as parameter.
* Use numeric OID
* User snmp_get_multi_oid
* Use snmp_get_multi_oid
* Remove default extension from vendor.
* Update test data.
* Renamed some mibs + changed oid in disco proc
* Updated json test data
* Update processor discovery to yaml
* fixed display bug when only one host had services
* small cleanup, renamed vars as laf change request
* message display wrongly removed, integrated back
* missed one changed line by #8030
* Restore vendor contents even if composer is not available
Leave the code in both places for safety.
* Instruct users to use the wrapper instead of straight up composer...
* Update librenms.logrotate
- Change path /opt/librenms/logs/*log so that logrotate catches http logs (access_log and error_log).
- Added SU parameter to rotate files with librenms user, as it was throwing insecure permissions error without it.
- Added CREATE parameter to create new log files as librenms user.
* Update librenms.logrotate
* device: Add XOS X480 and X650 desc and tests
Signed-off-by: Misha Komarovskiy <zombah@gmail.com>
* device: add XOS x480 and x650 test data
Signed-off-by: Misha Komarovskiy <zombah@gmail.com>
* Add missing hardware string in test data
* Update awplus to new processor discovery
Have no test data, so this needs testing.
* Updated YamlTest to include data as optional param for processors
* Share code between all mysql based authorizers
I plan to update the mysql password encryption and this will allow the code to be changed in a single location.
It also reduces a lot of duplication.
* Fix tests, I suspect reauthenticate will work for these...
Do not allow password updates for several authorizers
* Extract DiscoveryItem and move some things to better places.
Extract model class
Fix up model construction. I have problem with construction...
Makeshift model working. Switch constructor to factory. discover() and create()
Support legacy discovery.
Remove uneeded custom pollers
Remove netonix custom detection as we try ucd on all os now.
Add a few yaml procs. Fix a couple things.
More processor discovery conversions
Move Calix e7 to standard hrProcessorLoad, but it doesn't fully implement the HR-MIB, move things around to make it work.
Add a few yaml procs. Fix a couple things. Correct some stupid mib stuff.
Move more, drop php 5.3
Add netscaler which uses string indexes. Port fiberhome to yaml and use skip_values
More conversions. BroadcomProcessorUsage Trait
Serveriron and Ironware share some mibs. Create a common abstract os for them.
Add yaml support for mib specification in each data entry
Make legacy discover_processor() set 0 for hrDeviceIndex
Untangle Dell switch OS processors
Use use shared OS for groups if they don't have a specific group.
fix silly mib mistake
Make index optional
Move HR and UCD to Traits and out of Processor.
* forgot to update the fortiswitch index
* Make sgos and avaya-ers match the old index.
* fix comware test data
* fix merge errors
* fix dsm and remove pointless empty modules
* file not found exception is in the wrong place.
* Updated processor development docs
* initial
* commited correct version of f5-ltm-inc.php
* cleaned up comments
* cleaned up formating for travis
* split GTM discovery from LTM discovery module
* removed blank line at EOF
* updated per comments
* Update badge background for selected pagemenu item
This brings the background color in line with the user notification number badge background in the upper right corner.
The previous background was too bright for mono.css to properly read the white menu item text.
* CSS cache busting
Change last two digits to 42, because reasons.
Most of the time they should be polled as values anyway, but sometimes the mib might get loaded... If we set -Oe, then we always get the numeric value of the poller.
All sensors should get a numeric value anyway.
We still have to handle state sensors that return string values though.
* fix for syslog-ng config
I guess it didn't copy over correctly the last fix. It was reported to not be working I tested this and seems to be okay now at the time. It was showing 00:00 00 0000
* Update Syslog.md
* changed is_file to rrdtool_check_rrd_exists
* changed is_file to rrdtool_check_rrd_exists
* changed is_file to rrdtool_check_rrd_exists
* changed is_file to rrdtool_check_rrd_exists
* changed is_file to rrdtool_check_rrd_exists
